Associations between early lactation intramammary infections and udder health and performance during the first 180 days in milk in first-lactation organic dairy cows
Felipe Pena Mosca
8/18/2023
1 Exploratory Analysis
1.1 EA: Period prevalence IMI
table1(~IMI.cow + staph.aureus.IMI.cow + nas.IMI.cow + sch.IMI.cow + nas.non.chrom.IMI.cow + staph.epidermidis.IMI.cow + staph.haemolyticus.IMI.cow + staph.hominis.IMI.cow + staph.sciuri.IMI.cow + staph.xylosus.saprophyticus.IMI.cow + staph.spp.IMI.cow + strep.IMI.cow + strep.genus.IMI.cow + strep.uberis.IMI.cow + strep.dys.IMI.cow + strep.sp.IMI.cow + strep.like.IMI.cow + aerococcus.spp.IMI.cow + aero.vir.IMI.cow + aero.spp.IMI.cow + enterococcus.spp.IMI.cow + entero.cass.IMI.cow + entero.hyr.IMI.cow + entero.mun.IMI.cow + entero.spp.IMI.cow + entero.fae.IMI.cow + entero.sac.IMI.cow + lactococcus.spp.IMI.cow + lact.garv.IMI.cow + lact.lact.IMI.cow + lact.spp.IMI.cow + micro.spp.IMI.cow + gram.neg.IMI.cow + escherichia.spp.IMI.cow + e.coli.IMI.cow + e.spp.IMI.cow + kleb.IMI.cow + pseud.IMI.cow + citr.IMI.cow + enterobacter.genus.IMI.cow + ent.clo.IMI.cow+ent.spp.IMI.cow + gram.neg.org.IMI.cow + pantoea.IMI.cow + serratia.IMI.cow + other.IMI.cow + arthro.gand.IMI.cow + arthro.spp.IMI.cow + bacillus.IMI.cow + coryne.IMI.cow + gram.pos.cocci.IMI.cow + gram.pos.rod.IMI.cow + helcoccus.IMI.cow + trueperella.IMI.cow + yeast.IMI.cow |Site , data = df_dhi_cow)| A (N=49) |
B (N=116) |
C (N=105) |
E (N=63) |
Overall (N=333) |
|
|---|---|---|---|---|---|
| IMI.cow | |||||
| 0 | 26 (53.1%) | 10 (8.6%) | 5 (4.8%) | 0 (0%) | 41 (12.3%) |
| 1 | 23 (46.9%) | 106 (91.4%) | 100 (95.2%) | 63 (100%) | 292 (87.7%) |
| staph.aureus.IMI.cow | |||||
| 0 | 44 (89.8%) | 84 (72.4%) | 95 (90.5%) | 37 (58.7%) | 260 (78.1%) |
| 1 | 5 (10.2%) | 32 (27.6%) | 10 (9.5%) | 26 (41.3%) | 73 (21.9%) |
| nas.IMI.cow | |||||
| 0 | 35 (71.4%) | 46 (39.7%) | 62 (59.0%) | 23 (36.5%) | 166 (49.8%) |
| 1 | 14 (28.6%) | 70 (60.3%) | 43 (41.0%) | 40 (63.5%) | 167 (50.2%) |
| sch.IMI.cow | |||||
| 0 | 37 (75.5%) | 68 (58.6%) | 73 (69.5%) | 34 (54.0%) | 212 (63.7%) |
| 1 | 12 (24.5%) | 48 (41.4%) | 32 (30.5%) | 29 (46.0%) | 121 (36.3%) |
| nas.non.chrom.IMI.cow | |||||
| 0 | 46 (93.9%) | 82 (70.7%) | 92 (87.6%) | 40 (63.5%) | 260 (78.1%) |
| 1 | 3 (6.1%) | 34 (29.3%) | 13 (12.4%) | 23 (36.5%) | 73 (21.9%) |
| staph.epidermidis.IMI.cow | |||||
| 0 | 49 (100%) | 116 (100%) | 105 (100%) | 63 (100%) | 333 (100%) |
| 1 | 0 (0%) | 0 (0%) | 0 (0%) | 0 (0%) | 0 (0%) |
| staph.haemolyticus.IMI.cow | |||||
| 0 | 49 (100%) | 113 (97.4%) | 104 (99.0%) | 61 (96.8%) | 327 (98.2%) |
| 1 | 0 (0%) | 3 (2.6%) | 1 (1.0%) | 2 (3.2%) | 6 (1.8%) |
| staph.hominis.IMI.cow | |||||
| 0 | 49 (100%) | 116 (100%) | 105 (100%) | 62 (98.4%) | 332 (99.7%) |
| 1 | 0 (0%) | 0 (0%) | 0 (0%) | 1 (1.6%) | 1 (0.3%) |
| staph.sciuri.IMI.cow | |||||
| 0 | 49 (100%) | 116 (100%) | 105 (100%) | 59 (93.7%) | 329 (98.8%) |
| 1 | 0 (0%) | 0 (0%) | 0 (0%) | 4 (6.3%) | 4 (1.2%) |
| staph.xylosus.saprophyticus.IMI.cow | |||||
| 0 | 49 (100%) | 116 (100%) | 104 (99.0%) | 63 (100%) | 332 (99.7%) |
| 1 | 0 (0%) | 0 (0%) | 1 (1.0%) | 0 (0%) | 1 (0.3%) |
| staph.spp.IMI.cow | |||||
| 0 | 46 (93.9%) | 84 (72.4%) | 94 (89.5%) | 44 (69.8%) | 268 (80.5%) |
| 1 | 3 (6.1%) | 32 (27.6%) | 11 (10.5%) | 19 (30.2%) | 65 (19.5%) |
| strep.IMI.cow | |||||
| 0 | 45 (91.8%) | 81 (69.8%) | 61 (58.1%) | 25 (39.7%) | 212 (63.7%) |
| 1 | 4 (8.2%) | 35 (30.2%) | 44 (41.9%) | 38 (60.3%) | 121 (36.3%) |
| strep.genus.IMI.cow | |||||
| 0 | 48 (98.0%) | 87 (75.0%) | 96 (91.4%) | 35 (55.6%) | 266 (79.9%) |
| 1 | 1 (2.0%) | 29 (25.0%) | 9 (8.6%) | 28 (44.4%) | 67 (20.1%) |
| strep.uberis.IMI.cow | |||||
| 0 | 49 (100%) | 116 (100%) | 105 (100%) | 54 (85.7%) | 324 (97.3%) |
| 1 | 0 (0%) | 0 (0%) | 0 (0%) | 9 (14.3%) | 9 (2.7%) |
| strep.dys.IMI.cow | |||||
| 0 | 48 (98.0%) | 91 (78.4%) | 103 (98.1%) | 48 (76.2%) | 290 (87.1%) |
| 1 | 1 (2.0%) | 25 (21.6%) | 2 (1.9%) | 15 (23.8%) | 43 (12.9%) |
| strep.sp.IMI.cow | |||||
| 0 | 49 (100%) | 103 (88.8%) | 97 (92.4%) | 53 (84.1%) | 302 (90.7%) |
| 1 | 0 (0%) | 13 (11.2%) | 8 (7.6%) | 10 (15.9%) | 31 (9.3%) |
| strep.like.IMI.cow | |||||
| 0 | 46 (93.9%) | 110 (94.8%) | 66 (62.9%) | 51 (81.0%) | 273 (82.0%) |
| 1 | 3 (6.1%) | 6 (5.2%) | 39 (37.1%) | 12 (19.0%) | 60 (18.0%) |
| aerococcus.spp.IMI.cow | |||||
| 0 | 47 (95.9%) | 116 (100%) | 102 (97.1%) | 54 (85.7%) | 319 (95.8%) |
| 1 | 2 (4.1%) | 0 (0%) | 3 (2.9%) | 9 (14.3%) | 14 (4.2%) |
| aero.vir.IMI.cow | |||||
| 0 | 48 (98.0%) | 116 (100%) | 102 (97.1%) | 56 (88.9%) | 322 (96.7%) |
| 1 | 1 (2.0%) | 0 (0%) | 3 (2.9%) | 7 (11.1%) | 11 (3.3%) |
| aero.spp.IMI.cow | |||||
| 0 | 48 (98.0%) | 116 (100%) | 105 (100%) | 60 (95.2%) | 329 (98.8%) |
| 1 | 1 (2.0%) | 0 (0%) | 0 (0%) | 3 (4.8%) | 4 (1.2%) |
| enterococcus.spp.IMI.cow | |||||
| 0 | 49 (100%) | 111 (95.7%) | 72 (68.6%) | 62 (98.4%) | 294 (88.3%) |
| 1 | 0 (0%) | 5 (4.3%) | 33 (31.4%) | 1 (1.6%) | 39 (11.7%) |
| entero.cass.IMI.cow | |||||
| 0 | 49 (100%) | 115 (99.1%) | 94 (89.5%) | 63 (100%) | 321 (96.4%) |
| 1 | 0 (0%) | 1 (0.9%) | 11 (10.5%) | 0 (0%) | 12 (3.6%) |
| entero.hyr.IMI.cow | |||||
| 0 | 49 (100%) | 116 (100%) | 97 (92.4%) | 63 (100%) | 325 (97.6%) |
| 1 | 0 (0%) | 0 (0%) | 8 (7.6%) | 0 (0%) | 8 (2.4%) |
| entero.mun.IMI.cow | |||||
| 0 | 49 (100%) | 115 (99.1%) | 96 (91.4%) | 63 (100%) | 323 (97.0%) |
| 1 | 0 (0%) | 1 (0.9%) | 9 (8.6%) | 0 (0%) | 10 (3.0%) |
| entero.spp.IMI.cow | |||||
| 0 | 49 (100%) | 113 (97.4%) | 95 (90.5%) | 62 (98.4%) | 319 (95.8%) |
| 1 | 0 (0%) | 3 (2.6%) | 10 (9.5%) | 1 (1.6%) | 14 (4.2%) |
| entero.fae.IMI.cow | |||||
| 0 | 49 (100%) | 116 (100%) | 104 (99.0%) | 63 (100%) | 332 (99.7%) |
| 1 | 0 (0%) | 0 (0%) | 1 (1.0%) | 0 (0%) | 1 (0.3%) |
| entero.sac.IMI.cow | |||||
| 0 | 49 (100%) | 116 (100%) | 104 (99.0%) | 63 (100%) | 332 (99.7%) |
| 1 | 0 (0%) | 0 (0%) | 1 (1.0%) | 0 (0%) | 1 (0.3%) |
| lactococcus.spp.IMI.cow | |||||
| 0 | 48 (98.0%) | 116 (100%) | 101 (96.2%) | 63 (100%) | 328 (98.5%) |
| 1 | 1 (2.0%) | 0 (0%) | 4 (3.8%) | 0 (0%) | 5 (1.5%) |
| lact.garv.IMI.cow | |||||
| 0 | 48 (98.0%) | 116 (100%) | 103 (98.1%) | 63 (100%) | 330 (99.1%) |
| 1 | 1 (2.0%) | 0 (0%) | 2 (1.9%) | 0 (0%) | 3 (0.9%) |
| lact.lact.IMI.cow | |||||
| 0 | 49 (100%) | 116 (100%) | 104 (99.0%) | 63 (100%) | 332 (99.7%) |
| 1 | 0 (0%) | 0 (0%) | 1 (1.0%) | 0 (0%) | 1 (0.3%) |
| lact.spp.IMI.cow | |||||
| 0 | 49 (100%) | 116 (100%) | 104 (99.0%) | 63 (100%) | 332 (99.7%) |
| 1 | 0 (0%) | 0 (0%) | 1 (1.0%) | 0 (0%) | 1 (0.3%) |
| micro.spp.IMI.cow | |||||
| 0 | 49 (100%) | 115 (99.1%) | 104 (99.0%) | 61 (96.8%) | 329 (98.8%) |
| 1 | 0 (0%) | 1 (0.9%) | 1 (1.0%) | 2 (3.2%) | 4 (1.2%) |
| gram.neg.IMI.cow | |||||
| 0 | 49 (100%) | 110 (94.8%) | 83 (79.0%) | 60 (95.2%) | 302 (90.7%) |
| 1 | 0 (0%) | 6 (5.2%) | 22 (21.0%) | 3 (4.8%) | 31 (9.3%) |
| escherichia.spp.IMI.cow | |||||
| 0 | 49 (100%) | 112 (96.6%) | 101 (96.2%) | 61 (96.8%) | 323 (97.0%) |
| 1 | 0 (0%) | 4 (3.4%) | 4 (3.8%) | 2 (3.2%) | 10 (3.0%) |
| e.coli.IMI.cow | |||||
| 0 | 49 (100%) | 112 (96.6%) | 101 (96.2%) | 61 (96.8%) | 323 (97.0%) |
| 1 | 0 (0%) | 4 (3.4%) | 4 (3.8%) | 2 (3.2%) | 10 (3.0%) |
| e.spp.IMI.cow | |||||
| 0 | 49 (100%) | 115 (99.1%) | 105 (100%) | 62 (98.4%) | 331 (99.4%) |
| 1 | 0 (0%) | 1 (0.9%) | 0 (0%) | 1 (1.6%) | 2 (0.6%) |
| kleb.IMI.cow | |||||
| 0 | 49 (100%) | 116 (100%) | 102 (97.1%) | 63 (100%) | 330 (99.1%) |
| 1 | 0 (0%) | 0 (0%) | 3 (2.9%) | 0 (0%) | 3 (0.9%) |
| pseud.IMI.cow | |||||
| 0 | 49 (100%) | 116 (100%) | 102 (97.1%) | 63 (100%) | 330 (99.1%) |
| 1 | 0 (0%) | 0 (0%) | 3 (2.9%) | 0 (0%) | 3 (0.9%) |
| citr.IMI.cow | |||||
| 0 | 49 (100%) | 116 (100%) | 104 (99.0%) | 63 (100%) | 332 (99.7%) |
| 1 | 0 (0%) | 0 (0%) | 1 (1.0%) | 0 (0%) | 1 (0.3%) |
| enterobacter.genus.IMI.cow | |||||
| 0 | 49 (100%) | 116 (100%) | 104 (99.0%) | 62 (98.4%) | 331 (99.4%) |
| 1 | 0 (0%) | 0 (0%) | 1 (1.0%) | 1 (1.6%) | 2 (0.6%) |
| ent.clo.IMI.cow | |||||
| 0 | 49 (100%) | 116 (100%) | 105 (100%) | 62 (98.4%) | 332 (99.7%) |
| 1 | 0 (0%) | 0 (0%) | 0 (0%) | 1 (1.6%) | 1 (0.3%) |
| ent.spp.IMI.cow | |||||
| 0 | 49 (100%) | 116 (100%) | 104 (99.0%) | 63 (100%) | 332 (99.7%) |
| 1 | 0 (0%) | 0 (0%) | 1 (1.0%) | 0 (0%) | 1 (0.3%) |
| gram.neg.org.IMI.cow | |||||
| 0 | 49 (100%) | 115 (99.1%) | 102 (97.1%) | 63 (100%) | 329 (98.8%) |
| 1 | 0 (0%) | 1 (0.9%) | 3 (2.9%) | 0 (0%) | 4 (1.2%) |
| pantoea.IMI.cow | |||||
| 0 | 49 (100%) | 116 (100%) | 102 (97.1%) | 63 (100%) | 330 (99.1%) |
| 1 | 0 (0%) | 0 (0%) | 3 (2.9%) | 0 (0%) | 3 (0.9%) |
| serratia.IMI.cow | |||||
| 0 | 49 (100%) | 116 (100%) | 102 (97.1%) | 63 (100%) | 330 (99.1%) |
| 1 | 0 (0%) | 0 (0%) | 3 (2.9%) | 0 (0%) | 3 (0.9%) |
| other.IMI.cow | |||||
| 0 | 46 (93.9%) | 82 (70.7%) | 30 (28.6%) | 51 (81.0%) | 209 (62.8%) |
| 1 | 3 (6.1%) | 34 (29.3%) | 75 (71.4%) | 12 (19.0%) | 124 (37.2%) |
| arthro.gand.IMI.cow | |||||
| 0 | 49 (100%) | 115 (99.1%) | 104 (99.0%) | 63 (100%) | 331 (99.4%) |
| 1 | 0 (0%) | 1 (0.9%) | 1 (1.0%) | 0 (0%) | 2 (0.6%) |
| arthro.spp.IMI.cow | |||||
| 0 | 49 (100%) | 115 (99.1%) | 105 (100%) | 63 (100%) | 332 (99.7%) |
| 1 | 0 (0%) | 1 (0.9%) | 0 (0%) | 0 (0%) | 1 (0.3%) |
| bacillus.IMI.cow | |||||
| 0 | 49 (100%) | 98 (84.5%) | 54 (51.4%) | 53 (84.1%) | 254 (76.3%) |
| 1 | 0 (0%) | 18 (15.5%) | 51 (48.6%) | 10 (15.9%) | 79 (23.7%) |
| coryne.IMI.cow | |||||
| 0 | 48 (98.0%) | 106 (91.4%) | 63 (60.0%) | 63 (100%) | 280 (84.1%) |
| 1 | 1 (2.0%) | 10 (8.6%) | 42 (40.0%) | 0 (0%) | 53 (15.9%) |
| gram.pos.cocci.IMI.cow | |||||
| 0 | 49 (100%) | 113 (97.4%) | 94 (89.5%) | 63 (100%) | 319 (95.8%) |
| 1 | 0 (0%) | 3 (2.6%) | 11 (10.5%) | 0 (0%) | 14 (4.2%) |
| gram.pos.rod.IMI.cow | |||||
| 0 | 49 (100%) | 115 (99.1%) | 105 (100%) | 63 (100%) | 332 (99.7%) |
| 1 | 0 (0%) | 1 (0.9%) | 0 (0%) | 0 (0%) | 1 (0.3%) |
| helcoccus.IMI.cow | |||||
| 0 | 49 (100%) | 116 (100%) | 105 (100%) | 62 (98.4%) | 332 (99.7%) |
| 1 | 0 (0%) | 0 (0%) | 0 (0%) | 1 (1.6%) | 1 (0.3%) |
| trueperella.IMI.cow | |||||
| 0 | 48 (98.0%) | 115 (99.1%) | 105 (100%) | 63 (100%) | 331 (99.4%) |
| 1 | 1 (2.0%) | 1 (0.9%) | 0 (0%) | 0 (0%) | 2 (0.6%) |
| yeast.IMI.cow | |||||
| 0 | 49 (100%) | 114 (98.3%) | 105 (100%) | 63 (100%) | 331 (99.4%) |
| 1 | 0 (0%) | 2 (1.7%) | 0 (0%) | 0 (0%) | 2 (0.6%) |
1.2 EA: Period prevalence persistent IMI
table1(~staph.aureus.IMI.cow.p + nas.IMI.cow.p + sch.IMI.cow.p + nas.non.chrom.IMI.cow.p + staph.epidermidis.IMI.cow.p + staph.haemolyticus.IMI.cow.p + staph.hominis.IMI.cow.p + staph.sciuri.IMI.cow.p + staph.xylosus.saprophyticus.IMI.cow.p + staph.spp.IMI.cow.p + strep.IMI.cow.p + strep.genus.IMI.cow.p + strep.uberis.IMI.cow.p + strep.dys.IMI.cow.p + strep.sp.IMI.cow.p + strep.like.IMI.cow.p + aerococcus.spp.IMI.cow.p + aero.vir.IMI.cow.p + aero.spp.IMI.cow.p + enterococcus.spp.IMI.cow.p + entero.cass.IMI.cow.p + entero.hyr.IMI.cow.p + entero.mun.IMI.cow.p + entero.spp.IMI.cow.p + entero.fae.IMI.cow.p + entero.sac.IMI.cow.p +
lactococcus.spp.IMI.cow.p + lact.garv.IMI.cow.p + lact.lact.IMI.cow.p + lact.spp.IMI.cow.p +
micro.spp.IMI.cow.p + arthro.gand.IMI.cow.p + arthro.spp.IMI.cow.p + bacillus.IMI.cow.p + coryne.IMI.cow.p + gram.pos.cocci.IMI.cow.p + gram.pos.rod.IMI.cow.p + helcoccus.IMI.cow.p + trueperella.IMI.cow.p + yeast.IMI.cow.p |Site , data = df_dhi_cow)| A (N=49) |
B (N=116) |
C (N=105) |
E (N=63) |
Overall (N=333) |
|
|---|---|---|---|---|---|
| staph.aureus.IMI.cow.p | |||||
| 0 | 44 (89.8%) | 84 (72.4%) | 95 (90.5%) | 37 (58.7%) | 260 (78.1%) |
| 1 | 1 (2.0%) | 3 (2.6%) | 0 (0%) | 6 (9.5%) | 10 (3.0%) |
| 2 | 4 (8.2%) | 29 (25.0%) | 10 (9.5%) | 20 (31.7%) | 63 (18.9%) |
| nas.IMI.cow.p | |||||
| 0 | 35 (71.4%) | 46 (39.7%) | 62 (59.0%) | 23 (36.5%) | 166 (49.8%) |
| 1 | 4 (8.2%) | 23 (19.8%) | 28 (26.7%) | 18 (28.6%) | 73 (21.9%) |
| 2 | 10 (20.4%) | 47 (40.5%) | 15 (14.3%) | 22 (34.9%) | 94 (28.2%) |
| sch.IMI.cow.p | |||||
| 0 | 37 (75.5%) | 68 (58.6%) | 73 (69.5%) | 34 (54.0%) | 212 (63.7%) |
| 1 | 2 (4.1%) | 10 (8.6%) | 18 (17.1%) | 11 (17.5%) | 41 (12.3%) |
| 2 | 10 (20.4%) | 38 (32.8%) | 14 (13.3%) | 18 (28.6%) | 80 (24.0%) |
| nas.non.chrom.IMI.cow.p | |||||
| 0 | 46 (93.9%) | 82 (70.7%) | 92 (87.6%) | 40 (63.5%) | 260 (78.1%) |
| 1 | 3 (6.1%) | 20 (17.2%) | 12 (11.4%) | 16 (25.4%) | 51 (15.3%) |
| 2 | 0 (0%) | 14 (12.1%) | 1 (1.0%) | 7 (11.1%) | 22 (6.6%) |
| staph.epidermidis.IMI.cow.p | |||||
| 0 | 49 (100%) | 116 (100%) | 105 (100%) | 63 (100%) | 333 (100%) |
| 1 | 0 (0%) | 0 (0%) | 0 (0%) | 0 (0%) | 0 (0%) |
| staph.haemolyticus.IMI.cow.p | |||||
| 0 | 49 (100%) | 113 (97.4%) | 104 (99.0%) | 61 (96.8%) | 327 (98.2%) |
| 1 | 0 (0%) | 3 (2.6%) | 1 (1.0%) | 2 (3.2%) | 6 (1.8%) |
| staph.hominis.IMI.cow.p | |||||
| 0 | 49 (100%) | 116 (100%) | 105 (100%) | 62 (98.4%) | 332 (99.7%) |
| 1 | 0 (0%) | 0 (0%) | 0 (0%) | 1 (1.6%) | 1 (0.3%) |
| staph.sciuri.IMI.cow.p | |||||
| 0 | 49 (100%) | 116 (100%) | 105 (100%) | 59 (93.7%) | 329 (98.8%) |
| 1 | 0 (0%) | 0 (0%) | 0 (0%) | 4 (6.3%) | 4 (1.2%) |
| staph.xylosus.saprophyticus.IMI.cow.p | |||||
| 0 | 49 (100%) | 116 (100%) | 104 (99.0%) | 63 (100%) | 332 (99.7%) |
| 1 | 0 (0%) | 0 (0%) | 1 (1.0%) | 0 (0%) | 1 (0.3%) |
| 2 | 0 (0%) | 0 (0%) | 0 (0%) | 0 (0%) | 0 (0%) |
| staph.spp.IMI.cow.p | |||||
| 0 | 46 (93.9%) | 84 (72.4%) | 94 (89.5%) | 44 (69.8%) | 268 (80.5%) |
| 1 | 3 (6.1%) | 18 (15.5%) | 10 (9.5%) | 12 (19.0%) | 43 (12.9%) |
| 2 | 0 (0%) | 14 (12.1%) | 1 (1.0%) | 7 (11.1%) | 22 (6.6%) |
| strep.IMI.cow.p | |||||
| 0 | 45 (91.8%) | 81 (69.8%) | 61 (58.1%) | 25 (39.7%) | 212 (63.7%) |
| 1 | 3 (6.1%) | 19 (16.4%) | 40 (38.1%) | 24 (38.1%) | 86 (25.8%) |
| 2 | 1 (2.0%) | 16 (13.8%) | 4 (3.8%) | 14 (22.2%) | 35 (10.5%) |
| strep.genus.IMI.cow.p | |||||
| 0 | 48 (98.0%) | 87 (75.0%) | 96 (91.4%) | 35 (55.6%) | 266 (79.9%) |
| 1 | 0 (0%) | 14 (12.1%) | 8 (7.6%) | 14 (22.2%) | 36 (10.8%) |
| 2 | 1 (2.0%) | 15 (12.9%) | 1 (1.0%) | 14 (22.2%) | 31 (9.3%) |
| strep.uberis.IMI.cow.p | |||||
| 0 | 49 (100%) | 116 (100%) | 105 (100%) | 54 (85.7%) | 324 (97.3%) |
| 1 | 0 (0%) | 0 (0%) | 0 (0%) | 5 (7.9%) | 5 (1.5%) |
| 2 | 0 (0%) | 0 (0%) | 0 (0%) | 4 (6.3%) | 4 (1.2%) |
| strep.dys.IMI.cow.p | |||||
| 0 | 48 (98.0%) | 91 (78.4%) | 103 (98.1%) | 48 (76.2%) | 290 (87.1%) |
| 1 | 0 (0%) | 14 (12.1%) | 1 (1.0%) | 6 (9.5%) | 21 (6.3%) |
| 2 | 1 (2.0%) | 11 (9.5%) | 1 (1.0%) | 9 (14.3%) | 22 (6.6%) |
| strep.sp.IMI.cow.p | |||||
| 0 | 49 (100%) | 103 (88.8%) | 97 (92.4%) | 53 (84.1%) | 302 (90.7%) |
| 1 | 0 (0%) | 8 (6.9%) | 7 (6.7%) | 7 (11.1%) | 22 (6.6%) |
| 2 | 0 (0%) | 5 (4.3%) | 1 (1.0%) | 3 (4.8%) | 9 (2.7%) |
| strep.like.IMI.cow.p | |||||
| 0 | 46 (93.9%) | 110 (94.8%) | 66 (62.9%) | 51 (81.0%) | 273 (82.0%) |
| 1 | 3 (6.1%) | 4 (3.4%) | 35 (33.3%) | 10 (15.9%) | 52 (15.6%) |
| 2 | 0 (0%) | 2 (1.7%) | 4 (3.8%) | 2 (3.2%) | 8 (2.4%) |
| aerococcus.spp.IMI.cow.p | |||||
| 0 | 47 (95.9%) | 116 (100%) | 102 (97.1%) | 54 (85.7%) | 319 (95.8%) |
| 1 | 2 (4.1%) | 0 (0%) | 3 (2.9%) | 9 (14.3%) | 14 (4.2%) |
| 2 | 0 (0%) | 0 (0%) | 0 (0%) | 0 (0%) | 0 (0%) |
| aero.vir.IMI.cow.p | |||||
| 0 | 48 (98.0%) | 116 (100%) | 102 (97.1%) | 56 (88.9%) | 322 (96.7%) |
| 1 | 1 (2.0%) | 0 (0%) | 3 (2.9%) | 7 (11.1%) | 11 (3.3%) |
| 2 | 0 (0%) | 0 (0%) | 0 (0%) | 0 (0%) | 0 (0%) |
| aero.spp.IMI.cow.p | |||||
| 0 | 48 (98.0%) | 116 (100%) | 105 (100%) | 60 (95.2%) | 329 (98.8%) |
| 1 | 1 (2.0%) | 0 (0%) | 0 (0%) | 3 (4.8%) | 4 (1.2%) |
| enterococcus.spp.IMI.cow.p | |||||
| 0 | 49 (100%) | 111 (95.7%) | 72 (68.6%) | 62 (98.4%) | 294 (88.3%) |
| 1 | 0 (0%) | 4 (3.4%) | 30 (28.6%) | 1 (1.6%) | 35 (10.5%) |
| 2 | 0 (0%) | 1 (0.9%) | 3 (2.9%) | 0 (0%) | 4 (1.2%) |
| entero.cass.IMI.cow.p | |||||
| 0 | 49 (100%) | 115 (99.1%) | 94 (89.5%) | 63 (100%) | 321 (96.4%) |
| 1 | 0 (0%) | 1 (0.9%) | 10 (9.5%) | 0 (0%) | 11 (3.3%) |
| 2 | 0 (0%) | 0 (0%) | 1 (1.0%) | 0 (0%) | 1 (0.3%) |
| entero.hyr.IMI.cow.p | |||||
| 0 | 49 (100%) | 116 (100%) | 97 (92.4%) | 63 (100%) | 325 (97.6%) |
| 1 | 0 (0%) | 0 (0%) | 8 (7.6%) | 0 (0%) | 8 (2.4%) |
| entero.mun.IMI.cow.p | |||||
| 0 | 49 (100%) | 115 (99.1%) | 96 (91.4%) | 63 (100%) | 323 (97.0%) |
| 1 | 0 (0%) | 1 (0.9%) | 9 (8.6%) | 0 (0%) | 10 (3.0%) |
| entero.spp.IMI.cow.p | |||||
| 0 | 49 (100%) | 113 (97.4%) | 95 (90.5%) | 62 (98.4%) | 319 (95.8%) |
| 1 | 0 (0%) | 2 (1.7%) | 8 (7.6%) | 1 (1.6%) | 11 (3.3%) |
| 2 | 0 (0%) | 1 (0.9%) | 2 (1.9%) | 0 (0%) | 3 (0.9%) |
| entero.fae.IMI.cow.p | |||||
| 0 | 49 (100%) | 116 (100%) | 104 (99.0%) | 63 (100%) | 332 (99.7%) |
| 1 | 0 (0%) | 0 (0%) | 1 (1.0%) | 0 (0%) | 1 (0.3%) |
| entero.sac.IMI.cow.p | |||||
| 0 | 49 (100%) | 116 (100%) | 104 (99.0%) | 63 (100%) | 332 (99.7%) |
| 1 | 0 (0%) | 0 (0%) | 1 (1.0%) | 0 (0%) | 1 (0.3%) |
| lactococcus.spp.IMI.cow.p | |||||
| 0 | 48 (98.0%) | 116 (100%) | 101 (96.2%) | 63 (100%) | 328 (98.5%) |
| 1 | 1 (2.0%) | 0 (0%) | 4 (3.8%) | 0 (0%) | 5 (1.5%) |
| 2 | 0 (0%) | 0 (0%) | 0 (0%) | 0 (0%) | 0 (0%) |
| lact.garv.IMI.cow.p | |||||
| 0 | 48 (98.0%) | 116 (100%) | 103 (98.1%) | 63 (100%) | 330 (99.1%) |
| 1 | 1 (2.0%) | 0 (0%) | 2 (1.9%) | 0 (0%) | 3 (0.9%) |
| lact.lact.IMI.cow.p | |||||
| 0 | 49 (100%) | 116 (100%) | 104 (99.0%) | 63 (100%) | 332 (99.7%) |
| 1 | 0 (0%) | 0 (0%) | 1 (1.0%) | 0 (0%) | 1 (0.3%) |
| 2 | 0 (0%) | 0 (0%) | 0 (0%) | 0 (0%) | 0 (0%) |
| lact.spp.IMI.cow.p | |||||
| 0 | 49 (100%) | 116 (100%) | 104 (99.0%) | 63 (100%) | 332 (99.7%) |
| 1 | 0 (0%) | 0 (0%) | 1 (1.0%) | 0 (0%) | 1 (0.3%) |
| micro.spp.IMI.cow.p | |||||
| 0 | 49 (100%) | 115 (99.1%) | 104 (99.0%) | 61 (96.8%) | 329 (98.8%) |
| 1 | 0 (0%) | 1 (0.9%) | 1 (1.0%) | 2 (3.2%) | 4 (1.2%) |
| arthro.gand.IMI.cow.p | |||||
| 0 | 49 (100%) | 115 (99.1%) | 104 (99.0%) | 63 (100%) | 331 (99.4%) |
| 1 | 0 (0%) | 1 (0.9%) | 1 (1.0%) | 0 (0%) | 2 (0.6%) |
| arthro.spp.IMI.cow.p | |||||
| 0 | 49 (100%) | 115 (99.1%) | 105 (100%) | 63 (100%) | 332 (99.7%) |
| 1 | 0 (0%) | 1 (0.9%) | 0 (0%) | 0 (0%) | 1 (0.3%) |
| bacillus.IMI.cow.p | |||||
| 0 | 49 (100%) | 98 (84.5%) | 54 (51.4%) | 53 (84.1%) | 254 (76.3%) |
| 1 | 0 (0%) | 16 (13.8%) | 35 (33.3%) | 10 (15.9%) | 61 (18.3%) |
| 2 | 0 (0%) | 2 (1.7%) | 16 (15.2%) | 0 (0%) | 18 (5.4%) |
| coryne.IMI.cow.p | |||||
| 0 | 48 (98.0%) | 106 (91.4%) | 63 (60.0%) | 63 (100%) | 280 (84.1%) |
| 1 | 1 (2.0%) | 10 (8.6%) | 30 (28.6%) | 0 (0%) | 41 (12.3%) |
| 2 | 0 (0%) | 0 (0%) | 12 (11.4%) | 0 (0%) | 12 (3.6%) |
| gram.pos.cocci.IMI.cow.p | |||||
| 0 | 49 (100%) | 113 (97.4%) | 94 (89.5%) | 63 (100%) | 319 (95.8%) |
| 1 | 0 (0%) | 3 (2.6%) | 11 (10.5%) | 0 (0%) | 14 (4.2%) |
| gram.pos.rod.IMI.cow.p | |||||
| 0 | 49 (100%) | 115 (99.1%) | 105 (100%) | 63 (100%) | 332 (99.7%) |
| 1 | 0 (0%) | 1 (0.9%) | 0 (0%) | 0 (0%) | 1 (0.3%) |
| helcoccus.IMI.cow.p | |||||
| 0 | 49 (100%) | 116 (100%) | 105 (100%) | 62 (98.4%) | 332 (99.7%) |
| 1 | 0 (0%) | 0 (0%) | 0 (0%) | 1 (1.6%) | 1 (0.3%) |
| trueperella.IMI.cow.p | |||||
| 0 | 48 (98.0%) | 115 (99.1%) | 105 (100%) | 63 (100%) | 331 (99.4%) |
| 1 | 1 (2.0%) | 1 (0.9%) | 0 (0%) | 0 (0%) | 2 (0.6%) |
| yeast.IMI.cow.p | |||||
| 0 | 49 (100%) | 114 (98.3%) | 105 (100%) | 63 (100%) | 331 (99.4%) |
| 1 | 0 (0%) | 2 (1.7%) | 0 (0%) | 0 (0%) | 2 (0.6%) |
1.3 EA: Association between IMI and high SCC
table1(~as.factor(high_scc_cow)| staph.aureus.IMI.cow,data=df_survival)| 0 (N=260) |
1 (N=73) |
Overall (N=333) |
|
|---|---|---|---|
| as.factor(high_scc_cow) | |||
| 0 | 144 (55.4%) | 11 (15.1%) | 155 (46.5%) |
| 1 | 116 (44.6%) | 62 (84.9%) | 178 (53.5%) |
table1(~as.factor(high_scc_cow)| nas.IMI.cow,data=df_survival)| 0 (N=166) |
1 (N=167) |
Overall (N=333) |
|
|---|---|---|---|
| as.factor(high_scc_cow) | |||
| 0 | 86 (51.8%) | 69 (41.3%) | 155 (46.5%) |
| 1 | 80 (48.2%) | 98 (58.7%) | 178 (53.5%) |
table1(~as.factor(high_scc_cow)| sch.IMI.cow,data=df_survival)| 0 (N=212) |
1 (N=121) |
Overall (N=333) |
|
|---|---|---|---|
| as.factor(high_scc_cow) | |||
| 0 | 107 (50.5%) | 48 (39.7%) | 155 (46.5%) |
| 1 | 105 (49.5%) | 73 (60.3%) | 178 (53.5%) |
kmsurv <- survfit(survival ~ sch.IMI.cow, data = df_survival)
summary(kmsurv, times=c(1,2,3,4,5,6))## Call: survfit(formula = survival ~ sch.IMI.cow, data = df_survival)
##
## sch.IMI.cow=0
## time n.risk n.event survival std.err lower 95% CI upper 95% CI
## 1 212 68 0.679 0.0321 0.619 0.745
## 2 136 16 0.599 0.0339 0.536 0.670
## 3 108 4 0.577 0.0345 0.513 0.649
## 4 100 5 0.548 0.0351 0.484 0.621
## 5 93 3 0.531 0.0354 0.466 0.605
## 6 84 9 0.474 0.0363 0.408 0.551
##
## sch.IMI.cow=1
## time n.risk n.event survival std.err lower 95% CI upper 95% CI
## 1 121 32 0.736 0.0401 0.661 0.818
## 2 83 17 0.585 0.0456 0.502 0.681
## 3 63 6 0.529 0.0466 0.445 0.629
## 4 52 7 0.458 0.0475 0.374 0.561
## 5 45 3 0.427 0.0474 0.344 0.531
## 6 38 8 0.337 0.0469 0.257 0.443table1(~as.factor(high_scc_cow)| nas.non.chrom.IMI.cow,data=df_survival)| 0 (N=260) |
1 (N=73) |
Overall (N=333) |
|
|---|---|---|---|
| as.factor(high_scc_cow) | |||
| 0 | 126 (48.5%) | 29 (39.7%) | 155 (46.5%) |
| 1 | 134 (51.5%) | 44 (60.3%) | 178 (53.5%) |
table1(~as.factor(high_scc_cow)| staph.spp.IMI.cow,data=df_survival)| 0 (N=268) |
1 (N=65) |
Overall (N=333) |
|
|---|---|---|---|
| as.factor(high_scc_cow) | |||
| 0 | 129 (48.1%) | 26 (40.0%) | 155 (46.5%) |
| 1 | 139 (51.9%) | 39 (60.0%) | 178 (53.5%) |
table1(~as.factor(high_scc_cow)| strep.IMI.cow,data=df_survival)| 0 (N=212) |
1 (N=121) |
Overall (N=333) |
|
|---|---|---|---|
| as.factor(high_scc_cow) | |||
| 0 | 112 (52.8%) | 43 (35.5%) | 155 (46.5%) |
| 1 | 100 (47.2%) | 78 (64.5%) | 178 (53.5%) |
table1(~as.factor(high_scc_cow)| strep.genus.IMI.cow,data=df_survival)| 0 (N=266) |
1 (N=67) |
Overall (N=333) |
|
|---|---|---|---|
| as.factor(high_scc_cow) | |||
| 0 | 141 (53.0%) | 14 (20.9%) | 155 (46.5%) |
| 1 | 125 (47.0%) | 53 (79.1%) | 178 (53.5%) |
table1(~as.factor(high_scc_cow)| strep.dys.IMI.cow,data=df_survival)| 0 (N=290) |
1 (N=43) |
Overall (N=333) |
|
|---|---|---|---|
| as.factor(high_scc_cow) | |||
| 0 | 147 (50.7%) | 8 (18.6%) | 155 (46.5%) |
| 1 | 143 (49.3%) | 35 (81.4%) | 178 (53.5%) |
table1(~as.factor(high_scc_cow)| strep.sp.IMI.cow,data=df_survival)| 0 (N=302) |
1 (N=31) |
Overall (N=333) |
|
|---|---|---|---|
| as.factor(high_scc_cow) | |||
| 0 | 148 (49.0%) | 7 (22.6%) | 155 (46.5%) |
| 1 | 154 (51.0%) | 24 (77.4%) | 178 (53.5%) |
table1(~as.factor(high_scc_cow)| strep.like.IMI.cow,data=df_survival)| 0 (N=273) |
1 (N=60) |
Overall (N=333) |
|
|---|---|---|---|
| as.factor(high_scc_cow) | |||
| 0 | 123 (45.1%) | 32 (53.3%) | 155 (46.5%) |
| 1 | 150 (54.9%) | 28 (46.7%) | 178 (53.5%) |
table1(~as.factor(high_scc_cow)| as.factor(enterococcus.spp.IMI.cow),data=df_survival)| 0 (N=294) |
1 (N=39) |
Overall (N=333) |
|
|---|---|---|---|
| as.factor(high_scc_cow) | |||
| 0 | 133 (45.2%) | 22 (56.4%) | 155 (46.5%) |
| 1 | 161 (54.8%) | 17 (43.6%) | 178 (53.5%) |
table1(~as.factor(high_scc_cow)| as.factor(gram.neg.IMI.cow),data=df_survival)| 0 (N=302) |
1 (N=31) |
Overall (N=333) |
|
|---|---|---|---|
| as.factor(high_scc_cow) | |||
| 0 | 140 (46.4%) | 15 (48.4%) | 155 (46.5%) |
| 1 | 162 (53.6%) | 16 (51.6%) | 178 (53.5%) |
table1(~as.factor(high_scc_cow)| as.factor(other.IMI.cow),data=df_survival)| 0 (N=209) |
1 (N=124) |
Overall (N=333) |
|
|---|---|---|---|
| as.factor(high_scc_cow) | |||
| 0 | 79 (37.8%) | 76 (61.3%) | 155 (46.5%) |
| 1 | 130 (62.2%) | 48 (38.7%) | 178 (53.5%) |
table1(~as.factor(high_scc_cow)| as.factor(coryne.IMI.cow),data=df_survival)| 0 (N=280) |
1 (N=53) |
Overall (N=333) |
|
|---|---|---|---|
| as.factor(high_scc_cow) | |||
| 0 | 124 (44.3%) | 31 (58.5%) | 155 (46.5%) |
| 1 | 156 (55.7%) | 22 (41.5%) | 178 (53.5%) |
table1(~as.factor(high_scc_cow)| as.factor(bacillus.IMI.cow),data=df_survival)| 0 (N=254) |
1 (N=79) |
Overall (N=333) |
|
|---|---|---|---|
| as.factor(high_scc_cow) | |||
| 0 | 100 (39.4%) | 55 (69.6%) | 155 (46.5%) |
| 1 | 154 (60.6%) | 24 (30.4%) | 178 (53.5%) |
1.4 EA: Association between IMI and milk production
table1(~milk_kg+ as.factor(high_scc_cow)|Site:postpartum_sample,data=df.per.cow_dhi)## Warning in .table1.internal(x = x, labels = labels, groupspan = groupspan, :
## Table has 29 columns. Are you sure this is what you want?A |
B |
C |
E |
Overall |
|||||||||||||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 1 (N=47) |
2 (N=40) |
3 (N=40) |
5 (N=30) |
6 (N=27) |
1 (N=96) |
2 (N=115) |
3 (N=112) |
4 (N=108) |
5 (N=106) |
6 (N=107) |
1 (N=105) |
2 (N=94) |
3 (N=84) |
4 (N=80) |
5 (N=77) |
6 (N=69) |
1 (N=62) |
2 (N=61) |
3 (N=55) |
4 (N=54) |
5 (N=53) |
6 (N=52) |
1 (N=310) |
2 (N=310) |
3 (N=291) |
4 (N=242) |
5 (N=266) |
6 (N=255) |
|
| milk_kg | |||||||||||||||||||||||||||||
| Mean (SD) | 22.5 (5.13) | 25.5 (7.71) | 24.5 (6.36) | 19.2 (8.02) | 12.9 (5.66) | 29.2 (8.87) | 30.6 (6.88) | 30.7 (7.16) | 30.1 (6.08) | 31.5 (7.64) | 30.0 (6.49) | 26.1 (7.27) | 29.8 (4.74) | 28.9 (5.14) | 27.7 (4.71) | 26.2 (4.73) | 24.4 (5.12) | 17.6 (6.75) | 18.5 (6.70) | 15.9 (6.88) | 16.0 (8.09) | 14.5 (7.29) | 13.1 (5.76) | 24.8 (8.52) | 27.3 (7.88) | 26.5 (8.48) | 26.2 (8.30) | 25.2 (9.50) | 23.2 (9.30) |
| Median [Min, Max] | 22.2 [9.98, 32.7] | 24.3 [4.99, 59.0] | 24.0 [2.72, 36.3] | 17.5 [4.08, 35.8] | 12.7 [2.27, 25.4] | 28.8 [2.72, 54.4] | 30.4 [10.9, 46.3] | 30.4 [13.6, 54.9] | 29.9 [16.3, 54.4] | 29.9 [2.72, 49.0] | 29.9 [13.6, 54.4] | 26.3 [9.98, 41.7] | 29.9 [13.6, 39.9] | 29.0 [10.9, 38.1] | 28.1 [15.0, 38.1] | 26.3 [9.98, 35.4] | 24.9 [9.07, 35.4] | 17.2 [1.81, 28.6] | 17.7 [3.18, 30.8] | 14.5 [3.63, 30.8] | 13.6 [2.27, 35.4] | 12.7 [1.81, 30.4] | 12.2 [2.27, 27.2] | 25.9 [1.81, 54.4] | 27.2 [3.18, 59.0] | 27.7 [2.72, 54.9] | 27.7 [2.27, 54.4] | 26.3 [1.81, 49.0] | 25.4 [2.27, 54.4] |
| Missing | 0 (0%) | 0 (0%) | 0 (0%) | 0 (0%) | 0 (0%) | 0 (0%) | 0 (0%) | 0 (0%) | 0 (0%) | 0 (0%) | 0 (0%) | 0 (0%) | 0 (0%) | 0 (0%) | 0 (0%) | 0 (0%) | 0 (0%) | 1 (1.6%) | 0 (0%) | 0 (0%) | 0 (0%) | 0 (0%) | 0 (0%) | 1 (0.3%) | 0 (0%) | 0 (0%) | 0 (0%) | 0 (0%) | 0 (0%) |
| as.factor(high_scc_cow) | |||||||||||||||||||||||||||||
| 0 | 23 (48.9%) | 19 (47.5%) | 19 (47.5%) | 11 (36.7%) | 11 (40.7%) | 39 (40.6%) | 46 (40.0%) | 44 (39.3%) | 43 (39.8%) | 44 (41.5%) | 44 (41.1%) | 72 (68.6%) | 63 (67.0%) | 56 (66.7%) | 53 (66.3%) | 51 (66.2%) | 43 (62.3%) | 11 (17.7%) | 9 (14.8%) | 7 (12.7%) | 7 (13.0%) | 7 (13.2%) | 7 (13.5%) | 145 (46.8%) | 137 (44.2%) | 126 (43.3%) | 103 (42.6%) | 113 (42.5%) | 105 (41.2%) |
| 1 | 24 (51.1%) | 21 (52.5%) | 21 (52.5%) | 19 (63.3%) | 16 (59.3%) | 57 (59.4%) | 69 (60.0%) | 68 (60.7%) | 65 (60.2%) | 62 (58.5%) | 63 (58.9%) | 33 (31.4%) | 31 (33.0%) | 28 (33.3%) | 27 (33.8%) | 26 (33.8%) | 26 (37.7%) | 51 (82.3%) | 52 (85.2%) | 48 (87.3%) | 47 (87.0%) | 46 (86.8%) | 45 (86.5%) | 165 (53.2%) | 173 (55.8%) | 165 (56.7%) | 139 (57.4%) | 153 (57.5%) | 150 (58.8%) |
2 Confirmatory Analysis
2.1 CA: New high SCC
# ---- New subclinical mastitis ----
# Staph aureus
cox_1 <- coxph(survival~ staph.aureus.IMI.cow+cluster(Site), data=df_survival)
summary(cox_1)## Call:
## coxph(formula = survival ~ staph.aureus.IMI.cow, data = df_survival,
## cluster = Site)
##
## n= 333, number of events= 178
##
## coef exp(coef) se(coef) robust se z Pr(>|z|)
## staph.aureus.IMI.cow1 1.1692 3.2194 0.1608 0.1244 9.401 <2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## exp(coef) exp(-coef) lower .95 upper .95
## staph.aureus.IMI.cow1 3.219 0.3106 2.523 4.108
##
## Concordance= 0.621 (se = 0.029 )
## Likelihood ratio test= 45.83 on 1 df, p=1e-11
## Wald test = 88.37 on 1 df, p=<2e-16
## Score (logrank) test = 58.79 on 1 df, p=2e-14, Robust = 2.82 p=0.09
##
## (Note: the likelihood ratio and score tests assume independence of
## observations within a cluster, the Wald and robust score tests do not).cox.zph(cox_1)## chisq df p
## staph.aureus.IMI.cow 0.383 1 0.54
## GLOBAL 0.383 1 0.54cox_1_adj <- coxph(survival~ staph.aureus.IMI.cow+calving.nas+calving.sslo+cluster(Site), data=df_survival)
summary(cox_1_adj)## Call:
## coxph(formula = survival ~ staph.aureus.IMI.cow + calving.nas +
## calving.sslo, data = df_survival, cluster = Site)
##
## n= 333, number of events= 178
##
## coef exp(coef) se(coef) robust se z Pr(>|z|)
## staph.aureus.IMI.cow1 1.2086 3.3486 0.1626 0.1213 9.967 <2e-16 ***
## calving.nas1 0.4960 1.6421 0.1612 0.2246 2.208 0.0272 *
## calving.sslo1 0.8223 2.2756 0.1682 0.0744 11.052 <2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## exp(coef) exp(-coef) lower .95 upper .95
## staph.aureus.IMI.cow1 3.349 0.2986 2.640 4.247
## calving.nas1 1.642 0.6090 1.057 2.550
## calving.sslo1 2.276 0.4394 1.967 2.633
##
## Concordance= 0.703 (se = 0.037 )
## Likelihood ratio test= 73.35 on 3 df, p=8e-16
## Wald test = 15666 on 3 df, p=<2e-16
## Score (logrank) test = 89.33 on 3 df, p=<2e-16, Robust = 3.82 p=0.3
##
## (Note: the likelihood ratio and score tests assume independence of
## observations within a cluster, the Wald and robust score tests do not).cox.zph(cox_1_adj)## chisq df p
## staph.aureus.IMI.cow 0.795 1 0.37
## calving.nas 0.263 1 0.61
## calving.sslo 2.270 1 0.13
## GLOBAL 3.161 3 0.37# NAS general
cox_2 <- coxph(survival~ nas.IMI.cow+cluster(Site), data=df_survival)
summary(cox_2)## Call:
## coxph(formula = survival ~ nas.IMI.cow, data = df_survival, cluster = Site)
##
## n= 333, number of events= 178
##
## coef exp(coef) se(coef) robust se z Pr(>|z|)
## nas.IMI.cow1 0.2667 1.3057 0.1508 0.1478 1.804 0.0712 .
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## exp(coef) exp(-coef) lower .95 upper .95
## nas.IMI.cow1 1.306 0.7659 0.9772 1.745
##
## Concordance= 0.532 (se = 0.022 )
## Likelihood ratio test= 3.15 on 1 df, p=0.08
## Wald test = 3.25 on 1 df, p=0.07
## Score (logrank) test = 3.15 on 1 df, p=0.08, Robust = 1.71 p=0.2
##
## (Note: the likelihood ratio and score tests assume independence of
## observations within a cluster, the Wald and robust score tests do not).cox.zph(cox_2)## chisq df p
## nas.IMI.cow 1.39 1 0.24
## GLOBAL 1.39 1 0.24cox_2_adj <- coxph(survival~ nas.IMI.cow+calving.sau+calving.sslo+cluster(Site), data=df_survival)
summary(cox_2_adj)## Call:
## coxph(formula = survival ~ nas.IMI.cow + calving.sau + calving.sslo,
## data = df_survival, cluster = Site)
##
## n= 333, number of events= 178
##
## coef exp(coef) se(coef) robust se z Pr(>|z|)
## nas.IMI.cow1 0.29490 1.34300 0.15077 0.16425 1.795 0.0726 .
## calving.sau1 1.23236 3.42932 0.17045 0.14159 8.704 <2e-16 ***
## calving.sslo1 0.77331 2.16693 0.16672 0.04071 18.994 <2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## exp(coef) exp(-coef) lower .95 upper .95
## nas.IMI.cow1 1.343 0.7446 0.9733 1.853
## calving.sau1 3.429 0.2916 2.5983 4.526
## calving.sslo1 2.167 0.4615 2.0007 2.347
##
## Concordance= 0.703 (se = 0.037 )
## Likelihood ratio test= 66.95 on 3 df, p=2e-14
## Wald test = 744.7 on 3 df, p=<2e-16
## Score (logrank) test = 87.01 on 3 df, p=<2e-16, Robust = 3.97 p=0.3
##
## (Note: the likelihood ratio and score tests assume independence of
## observations within a cluster, the Wald and robust score tests do not).cox.zph(cox_2_adj)## chisq df p
## nas.IMI.cow 0.6121 1 0.434
## calving.sau 0.0186 1 0.892
## calving.sslo 3.3031 1 0.069
## GLOBAL 3.9042 3 0.272# Staph chromogenes
cox_3 <- coxph(survival~ sch.IMI.cow+cluster(Site), data=df_survival)
summary(cox_3)## Call:
## coxph(formula = survival ~ sch.IMI.cow, data = df_survival, cluster = Site)
##
## n= 333, number of events= 178
##
## coef exp(coef) se(coef) robust se z Pr(>|z|)
## sch.IMI.cow1 0.2158 1.2408 0.1526 0.1764 1.223 0.221
##
## exp(coef) exp(-coef) lower .95 upper .95
## sch.IMI.cow1 1.241 0.8059 0.8781 1.753
##
## Concordance= 0.512 (se = 0.025 )
## Likelihood ratio test= 1.97 on 1 df, p=0.2
## Wald test = 1.5 on 1 df, p=0.2
## Score (logrank) test = 2.01 on 1 df, p=0.2, Robust = 1.24 p=0.3
##
## (Note: the likelihood ratio and score tests assume independence of
## observations within a cluster, the Wald and robust score tests do not).cox.zph(cox_3)## chisq df p
## sch.IMI.cow 10.1 1 0.0015
## GLOBAL 10.1 1 0.0015cox_3_adj <- coxph(survival~ sch.IMI.cow+calving.sau+calving.sslo+cluster(Site), data=df_survival)
summary(cox_3_adj)## Call:
## coxph(formula = survival ~ sch.IMI.cow + calving.sau + calving.sslo,
## data = df_survival, cluster = Site)
##
## n= 333, number of events= 178
##
## coef exp(coef) se(coef) robust se z Pr(>|z|)
## sch.IMI.cow1 0.21054 1.23435 0.15263 0.18246 1.154 0.249
## calving.sau1 1.22635 3.40876 0.17065 0.13465 9.108 <2e-16 ***
## calving.sslo1 0.76587 2.15087 0.16665 0.05134 14.919 <2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## exp(coef) exp(-coef) lower .95 upper .95
## sch.IMI.cow1 1.234 0.8101 0.8632 1.765
## calving.sau1 3.409 0.2934 2.6181 4.438
## calving.sslo1 2.151 0.4649 1.9450 2.379
##
## Concordance= 0.693 (se = 0.038 )
## Likelihood ratio test= 64.98 on 3 df, p=5e-14
## Wald test = 2868 on 3 df, p=<2e-16
## Score (logrank) test = 84.97 on 3 df, p=<2e-16, Robust = 3.79 p=0.3
##
## (Note: the likelihood ratio and score tests assume independence of
## observations within a cluster, the Wald and robust score tests do not).cox.zph(cox_3_adj)## chisq df p
## sch.IMI.cow 8.0589 1 0.0045
## calving.sau 0.0377 1 0.8461
## calving.sslo 3.2923 1 0.0696
## GLOBAL 11.5058 3 0.0093# Proportional hazards assumption not met
cox_3_adj_stratified <- coxph(survival~ sch.IMI.cow*ps_sample_3_to_6 + calving.sau+calving.sslo+cluster(Site), data=df_survival)## Warning in coxph.fit(X, Y, istrat, offset, init, control, weights = weights, :
## Loglik converged before variable 2 ; coefficient may be infinite.summary(cox_3_adj_stratified)## Call:
## coxph(formula = survival ~ sch.IMI.cow + ps_sample_3_to_6 + calving.sau +
## calving.sslo + sch.IMI.cow:ps_sample_3_to_6, data = df_survival,
## cluster = Site)
##
## n= 333, number of events= 178
##
## coef exp(coef) se(coef) robust se
## sch.IMI.cow1 -1.428e-01 8.669e-01 1.737e-01 1.293e-01
## ps_sample_3_to_6 -2.153e+01 4.445e-10 2.156e+03 7.008e-01
## calving.sau1 5.461e-01 1.726e+00 1.723e-01 1.525e-01
## calving.sslo1 7.417e-01 2.100e+00 1.706e-01 9.965e-02
## sch.IMI.cow1:ps_sample_3_to_6 9.522e-01 2.591e+00 3.808e-01 2.165e-01
## z Pr(>|z|)
## sch.IMI.cow1 -1.105 0.269187
## ps_sample_3_to_6 -30.726 < 2e-16 ***
## calving.sau1 3.581 0.000342 ***
## calving.sslo1 7.443 9.83e-14 ***
## sch.IMI.cow1:ps_sample_3_to_6 4.399 1.09e-05 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## exp(coef) exp(-coef) lower .95 upper .95
## sch.IMI.cow1 8.669e-01 1.154e+00 6.729e-01 1.117e+00
## ps_sample_3_to_6 4.445e-10 2.250e+09 1.125e-10 1.756e-09
## calving.sau1 1.726e+00 5.792e-01 1.280e+00 2.328e+00
## calving.sslo1 2.100e+00 4.763e-01 1.727e+00 2.552e+00
## sch.IMI.cow1:ps_sample_3_to_6 2.591e+00 3.859e-01 1.695e+00 3.961e+00
##
## Concordance= 0.874 (se = 0.01 )
## Likelihood ratio test= 333.1 on 5 df, p=<2e-16
## Wald test = 2351 on 5 df, p=<2e-16
## Score (logrank) test = 281.6 on 5 df, p=<2e-16, Robust = 4 p=0.5
##
## (Note: the likelihood ratio and score tests assume independence of
## observations within a cluster, the Wald and robust score tests do not).confint(emmeans(cox_3_adj_stratified,revpairwise~sch.IMI.cow|ps_sample_3_to_6,type="response")$contrasts)## ps_sample_3_to_6 = 0:
## contrast ratio SE df asymp.LCL asymp.UCL
## sch.IMI.cow1 / sch.IMI.cow0 0.867 0.112 Inf 0.673 1.12
##
## ps_sample_3_to_6 = 1:
## contrast ratio SE df asymp.LCL asymp.UCL
## sch.IMI.cow1 / sch.IMI.cow0 2.247 0.400 Inf 1.584 3.19
##
## Results are averaged over the levels of: calving.sau, calving.sslo
## Confidence level used: 0.95
## Intervals are back-transformed from the log scale# NAS non-chromogenes
cox_4 <- coxph(survival~ nas.non.chrom.IMI.cow+cluster(Site), data=df_survival)
summary(cox_4)## Call:
## coxph(formula = survival ~ nas.non.chrom.IMI.cow, data = df_survival,
## cluster = Site)
##
## n= 333, number of events= 178
##
## coef exp(coef) se(coef) robust se z Pr(>|z|)
## nas.non.chrom.IMI.cow1 0.29314 1.34064 0.17386 0.07409 3.957 7.6e-05 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## exp(coef) exp(-coef) lower .95 upper .95
## nas.non.chrom.IMI.cow1 1.341 0.7459 1.159 1.55
##
## Concordance= 0.538 (se = 0.012 )
## Likelihood ratio test= 2.7 on 1 df, p=0.1
## Wald test = 15.66 on 1 df, p=8e-05
## Score (logrank) test = 2.86 on 1 df, p=0.09, Robust = 1.66 p=0.2
##
## (Note: the likelihood ratio and score tests assume independence of
## observations within a cluster, the Wald and robust score tests do not).cox.zph(cox_4)## chisq df p
## nas.non.chrom.IMI.cow 2.28 1 0.13
## GLOBAL 2.28 1 0.13cox_4_adj <- coxph(survival~ nas.non.chrom.IMI.cow+calving.sau+calving.sslo+cluster(Site), data=df_survival)
summary(cox_4_adj)## Call:
## coxph(formula = survival ~ nas.non.chrom.IMI.cow + calving.sau +
## calving.sslo, data = df_survival, cluster = Site)
##
## n= 333, number of events= 178
##
## coef exp(coef) se(coef) robust se z Pr(>|z|)
## nas.non.chrom.IMI.cow1 0.35978 1.43301 0.17458 0.15243 2.360 0.0183 *
## calving.sau1 1.26138 3.53029 0.17141 0.13125 9.611 <2e-16 ***
## calving.sslo1 0.76120 2.14084 0.16672 0.04429 17.187 <2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## exp(coef) exp(-coef) lower .95 upper .95
## nas.non.chrom.IMI.cow1 1.433 0.6978 1.063 1.932
## calving.sau1 3.530 0.2833 2.730 4.566
## calving.sslo1 2.141 0.4671 1.963 2.335
##
## Concordance= 0.695 (se = 0.04 )
## Likelihood ratio test= 67.09 on 3 df, p=2e-14
## Wald test = 574.7 on 3 df, p=<2e-16
## Score (logrank) test = 86.77 on 3 df, p=<2e-16, Robust = 3.77 p=0.3
##
## (Note: the likelihood ratio and score tests assume independence of
## observations within a cluster, the Wald and robust score tests do not).cox.zph(cox_4_adj)## chisq df p
## nas.non.chrom.IMI.cow 3.310 1 0.069
## calving.sau 0.051 1 0.821
## calving.sslo 3.054 1 0.081
## GLOBAL 6.343 3 0.096# Staph sp.
cox_5 <- coxph(survival~ staph.spp.IMI.cow+cluster(Site), data=df_survival)
summary(cox_5)## Call:
## coxph(formula = survival ~ staph.spp.IMI.cow, data = df_survival,
## cluster = Site)
##
## n= 333, number of events= 178
##
## coef exp(coef) se(coef) robust se z Pr(>|z|)
## staph.spp.IMI.cow1 0.2769 1.3190 0.1813 0.1018 2.721 0.00652 **
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## exp(coef) exp(-coef) lower .95 upper .95
## staph.spp.IMI.cow1 1.319 0.7581 1.08 1.61
##
## Concordance= 0.533 (se = 0.011 )
## Likelihood ratio test= 2.21 on 1 df, p=0.1
## Wald test = 7.4 on 1 df, p=0.007
## Score (logrank) test = 2.35 on 1 df, p=0.1, Robust = 1.51 p=0.2
##
## (Note: the likelihood ratio and score tests assume independence of
## observations within a cluster, the Wald and robust score tests do not).cox.zph(cox_5)## chisq df p
## staph.spp.IMI.cow 1.98 1 0.16
## GLOBAL 1.98 1 0.16cox_5_adj <- coxph(survival~ staph.spp.IMI.cow+calving.sau+calving.sslo+cluster(Site), data=df_survival)
summary(cox_5_adj)## Call:
## coxph(formula = survival ~ staph.spp.IMI.cow + calving.sau +
## calving.sslo, data = df_survival, cluster = Site)
##
## n= 333, number of events= 178
##
## coef exp(coef) se(coef) robust se z Pr(>|z|)
## staph.spp.IMI.cow1 0.3211 1.3787 0.1818 0.1465 2.192 0.0284 *
## calving.sau1 1.2521 3.4975 0.1711 0.1288 9.721 <2e-16 ***
## calving.sslo1 0.7585 2.1351 0.1668 0.0410 18.499 <2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## exp(coef) exp(-coef) lower .95 upper .95
## staph.spp.IMI.cow1 1.379 0.7253 1.035 1.837
## calving.sau1 3.498 0.2859 2.717 4.502
## calving.sslo1 2.135 0.4684 1.970 2.314
##
## Concordance= 0.692 (se = 0.037 )
## Likelihood ratio test= 66.04 on 3 df, p=3e-14
## Wald test = 537 on 3 df, p=<2e-16
## Score (logrank) test = 85.78 on 3 df, p=<2e-16, Robust = 3.72 p=0.3
##
## (Note: the likelihood ratio and score tests assume independence of
## observations within a cluster, the Wald and robust score tests do not).cox.zph(cox_5_adj)## chisq df p
## staph.spp.IMI.cow 2.9827 1 0.084
## calving.sau 0.0524 1 0.819
## calving.sslo 3.0820 1 0.079
## GLOBAL 5.9580 3 0.114# SSLO
cox_6 <- coxph(survival~ strep.IMI.cow+cluster(Site), data=df_survival)
summary(cox_6)## Call:
## coxph(formula = survival ~ strep.IMI.cow, data = df_survival,
## cluster = Site)
##
## n= 333, number of events= 178
##
## coef exp(coef) se(coef) robust se z Pr(>|z|)
## strep.IMI.cow1 0.6004 1.8228 0.1517 0.1093 5.495 3.91e-08 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## exp(coef) exp(-coef) lower .95 upper .95
## strep.IMI.cow1 1.823 0.5486 1.471 2.258
##
## Concordance= 0.595 (se = 0.023 )
## Likelihood ratio test= 15.09 on 1 df, p=1e-04
## Wald test = 30.19 on 1 df, p=4e-08
## Score (logrank) test = 16.14 on 1 df, p=6e-05, Robust = 3.11 p=0.08
##
## (Note: the likelihood ratio and score tests assume independence of
## observations within a cluster, the Wald and robust score tests do not).cox.zph(cox_6)## chisq df p
## strep.IMI.cow 5.65 1 0.017
## GLOBAL 5.65 1 0.017cox_6_adj <- coxph(survival~ strep.IMI.cow+calving.sau+calving.nas+cluster(Site), data=df_survival)
summary(cox_6_adj)## Call:
## coxph(formula = survival ~ strep.IMI.cow + calving.sau + calving.nas,
## data = df_survival, cluster = Site)
##
## n= 333, number of events= 178
##
## coef exp(coef) se(coef) robust se z Pr(>|z|)
## strep.IMI.cow1 0.62058 1.86001 0.15361 0.03288 18.874 <2e-16 ***
## calving.sau1 1.23649 3.44351 0.17136 0.09533 12.970 <2e-16 ***
## calving.nas1 0.42234 1.52552 0.15999 0.21303 1.982 0.0474 *
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## exp(coef) exp(-coef) lower .95 upper .95
## strep.IMI.cow1 1.860 0.5376 1.744 1.984
## calving.sau1 3.444 0.2904 2.857 4.151
## calving.nas1 1.526 0.6555 1.005 2.316
##
## Concordance= 0.695 (se = 0.034 )
## Likelihood ratio test= 64.95 on 3 df, p=5e-14
## Wald test = 36374 on 3 df, p=<2e-16
## Score (logrank) test = 82.72 on 3 df, p=<2e-16, Robust = 3.19 p=0.4
##
## (Note: the likelihood ratio and score tests assume independence of
## observations within a cluster, the Wald and robust score tests do not).cox.zph(cox_6_adj)## chisq df p
## strep.IMI.cow 3.565 1 0.059
## calving.sau 0.199 1 0.656
## calving.nas 0.446 1 0.504
## GLOBAL 3.869 3 0.276# Strep species
cox_7 <- coxph(survival~ strep.genus.IMI.cow+cluster(Site), data=df_survival)
summary(cox_7)## Call:
## coxph(formula = survival ~ strep.genus.IMI.cow, data = df_survival,
## cluster = Site)
##
## n= 333, number of events= 178
##
## coef exp(coef) se(coef) robust se z Pr(>|z|)
## strep.genus.IMI.cow1 0.9987 2.7148 0.1655 0.1110 9 <2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## exp(coef) exp(-coef) lower .95 upper .95
## strep.genus.IMI.cow1 2.715 0.3683 2.184 3.374
##
## Concordance= 0.606 (se = 0.029 )
## Likelihood ratio test= 31.26 on 1 df, p=2e-08
## Wald test = 81.01 on 1 df, p=<2e-16
## Score (logrank) test = 39.45 on 1 df, p=3e-10, Robust = 2.71 p=0.1
##
## (Note: the likelihood ratio and score tests assume independence of
## observations within a cluster, the Wald and robust score tests do not).cox.zph(cox_7)## chisq df p
## strep.genus.IMI.cow 3.55 1 0.059
## GLOBAL 3.55 1 0.059cox_7_adj <- coxph(survival~ strep.genus.IMI.cow+calving.sau+calving.nas+cluster(Site), data=df_survival)
summary(cox_7_adj)## Call:
## coxph(formula = survival ~ strep.genus.IMI.cow + calving.sau +
## calving.nas, data = df_survival, cluster = Site)
##
## n= 333, number of events= 178
##
## coef exp(coef) se(coef) robust se z Pr(>|z|)
## strep.genus.IMI.cow1 0.80967 2.24717 0.17040 0.03037 26.665 <2e-16 ***
## calving.sau1 1.10689 3.02493 0.17558 0.07323 15.116 <2e-16 ***
## calving.nas1 0.36051 1.43406 0.15887 0.18801 1.918 0.0552 .
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## exp(coef) exp(-coef) lower .95 upper .95
## strep.genus.IMI.cow1 2.247 0.4450 2.1173 2.385
## calving.sau1 3.025 0.3306 2.6205 3.492
## calving.nas1 1.434 0.6973 0.9921 2.073
##
## Concordance= 0.703 (se = 0.042 )
## Likelihood ratio test= 69.43 on 3 df, p=6e-15
## Wald test = 1216 on 3 df, p=<2e-16
## Score (logrank) test = 92.07 on 3 df, p=<2e-16, Robust = 2.76 p=0.4
##
## (Note: the likelihood ratio and score tests assume independence of
## observations within a cluster, the Wald and robust score tests do not).cox.zph(cox_7_adj)## chisq df p
## strep.genus.IMI.cow 2.057 1 0.15
## calving.sau 0.210 1 0.65
## calving.nas 0.185 1 0.67
## GLOBAL 2.690 3 0.44# Strep dysgalactiae
cox_8 <- coxph(survival~ strep.dys.IMI.cow+cluster(Site), data=df_survival)
summary(cox_8)## Call:
## coxph(formula = survival ~ strep.dys.IMI.cow, data = df_survival,
## cluster = Site)
##
## n= 333, number of events= 178
##
## coef exp(coef) se(coef) robust se z Pr(>|z|)
## strep.dys.IMI.cow1 0.9069 2.4766 0.1898 0.1904 4.763 1.91e-06 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## exp(coef) exp(-coef) lower .95 upper .95
## strep.dys.IMI.cow1 2.477 0.4038 1.705 3.597
##
## Concordance= 0.567 (se = 0.01 )
## Likelihood ratio test= 18.93 on 1 df, p=1e-05
## Wald test = 22.68 on 1 df, p=2e-06
## Score (logrank) test = 24.43 on 1 df, p=8e-07, Robust = 2.91 p=0.09
##
## (Note: the likelihood ratio and score tests assume independence of
## observations within a cluster, the Wald and robust score tests do not).cox.zph(cox_8)## chisq df p
## strep.dys.IMI.cow 0.217 1 0.64
## GLOBAL 0.217 1 0.64cox_8_adj <- coxph(survival~ strep.dys.IMI.cow+calving.sau+calving.nas+cluster(Site), data=df_survival)
summary(cox_8_adj)## Call:
## coxph(formula = survival ~ strep.dys.IMI.cow + calving.sau +
## calving.nas, data = df_survival, cluster = Site)
##
## n= 333, number of events= 178
##
## coef exp(coef) se(coef) robust se z Pr(>|z|)
## strep.dys.IMI.cow1 0.73096 2.07708 0.19250 0.09595 7.618 2.57e-14 ***
## calving.sau1 1.17691 3.24434 0.17289 0.11026 10.674 < 2e-16 ***
## calving.nas1 0.34139 1.40690 0.15885 0.17731 1.925 0.0542 .
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## exp(coef) exp(-coef) lower .95 upper .95
## strep.dys.IMI.cow1 2.077 0.4814 1.7210 2.507
## calving.sau1 3.244 0.3082 2.6138 4.027
## calving.nas1 1.407 0.7108 0.9939 1.992
##
## Concordance= 0.691 (se = 0.041 )
## Likelihood ratio test= 61.71 on 3 df, p=3e-13
## Wald test = 395.2 on 3 df, p=<2e-16
## Score (logrank) test = 82.96 on 3 df, p=<2e-16, Robust = 3.27 p=0.4
##
## (Note: the likelihood ratio and score tests assume independence of
## observations within a cluster, the Wald and robust score tests do not).cox.zph(cox_8_adj)## chisq df p
## strep.dys.IMI.cow 0.0873 1 0.77
## calving.sau 0.0428 1 0.84
## calving.nas 0.2098 1 0.65
## GLOBAL 0.3696 3 0.95# Unspeciated Strep
cox_9 <- coxph(survival~ strep.sp.IMI.cow+cluster(Site), data=df_survival)
summary(cox_9)## Call:
## coxph(formula = survival ~ strep.sp.IMI.cow, data = df_survival,
## cluster = Site)
##
## n= 333, number of events= 178
##
## coef exp(coef) se(coef) robust se z Pr(>|z|)
## strep.sp.IMI.cow1 0.7206 2.0557 0.2199 0.0704 10.24 <2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## exp(coef) exp(-coef) lower .95 upper .95
## strep.sp.IMI.cow1 2.056 0.4864 1.791 2.36
##
## Concordance= 0.54 (se = 0.01 )
## Likelihood ratio test= 9.02 on 1 df, p=0.003
## Wald test = 104.8 on 1 df, p=<2e-16
## Score (logrank) test = 11.21 on 1 df, p=8e-04, Robust = 2.34 p=0.1
##
## (Note: the likelihood ratio and score tests assume independence of
## observations within a cluster, the Wald and robust score tests do not).cox.zph(cox_9)## chisq df p
## strep.sp.IMI.cow 0.352 1 0.55
## GLOBAL 0.352 1 0.55cox_9_adj <- coxph(survival~ strep.sp.IMI.cow+calving.sau+calving.nas+cluster(Site), data=df_survival)
summary(cox_9_adj)## Call:
## coxph(formula = survival ~ strep.sp.IMI.cow + calving.sau + calving.nas,
## data = df_survival, cluster = Site)
##
## n= 333, number of events= 178
##
## coef exp(coef) se(coef) robust se z Pr(>|z|)
## strep.sp.IMI.cow1 0.5675 1.7638 0.2217 0.1741 3.260 0.00112 **
## calving.sau1 1.2179 3.3802 0.1718 0.1298 9.381 < 2e-16 ***
## calving.nas1 0.3628 1.4373 0.1589 0.1946 1.864 0.06230 .
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## exp(coef) exp(-coef) lower .95 upper .95
## strep.sp.IMI.cow1 1.764 0.5670 1.2539 2.481
## calving.sau1 3.380 0.2958 2.6207 4.360
## calving.nas1 1.437 0.6957 0.9815 2.105
##
## Concordance= 0.675 (se = 0.037 )
## Likelihood ratio test= 54.94 on 3 df, p=7e-12
## Wald test = 247.1 on 3 df, p=<2e-16
## Score (logrank) test = 72.99 on 3 df, p=1e-15, Robust = 2.76 p=0.4
##
## (Note: the likelihood ratio and score tests assume independence of
## observations within a cluster, the Wald and robust score tests do not).cox.zph(cox_9_adj)## chisq df p
## strep.sp.IMI.cow 0.3137 1 0.58
## calving.sau 0.0138 1 0.91
## calving.nas 0.1879 1 0.66
## GLOBAL 0.5330 3 0.91# Strep-like organisms
cox_10 <- coxph(survival~ strep.like.IMI.cow+cluster(Site), data=df_survival)
summary(cox_10)## Call:
## coxph(formula = survival ~ strep.like.IMI.cow, data = df_survival,
## cluster = Site)
##
## n= 333, number of events= 178
##
## coef exp(coef) se(coef) robust se z Pr(>|z|)
## strep.like.IMI.cow1 -0.1435 0.8663 0.2059 0.2054 -0.699 0.485
##
## exp(coef) exp(-coef) lower .95 upper .95
## strep.like.IMI.cow1 0.8663 1.154 0.5792 1.296
##
## Concordance= 0.508 (se = 0.015 )
## Likelihood ratio test= 0.5 on 1 df, p=0.5
## Wald test = 0.49 on 1 df, p=0.5
## Score (logrank) test = 0.49 on 1 df, p=0.5, Robust = 0.49 p=0.5
##
## (Note: the likelihood ratio and score tests assume independence of
## observations within a cluster, the Wald and robust score tests do not).cox.zph(cox_10)## chisq df p
## strep.like.IMI.cow 0.515 1 0.47
## GLOBAL 0.515 1 0.47cox_10_adj <- coxph(survival~ strep.like.IMI.cow+calving.sau+calving.nas+cluster(Site), data=df_survival)
summary(cox_10_adj)## Call:
## coxph(formula = survival ~ strep.like.IMI.cow + calving.sau +
## calving.nas, data = df_survival, cluster = Site)
##
## n= 333, number of events= 178
##
## coef exp(coef) se(coef) robust se z Pr(>|z|)
## strep.like.IMI.cow1 0.07462 1.07748 0.21001 0.20903 0.357 0.721
## calving.sau1 1.27058 3.56292 0.17214 0.14324 8.870 <2e-16 ***
## calving.nas1 0.35983 1.43309 0.16051 0.22385 1.607 0.108
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## exp(coef) exp(-coef) lower .95 upper .95
## strep.like.IMI.cow1 1.077 0.9281 0.7153 1.623
## calving.sau1 3.563 0.2807 2.6908 4.718
## calving.nas1 1.433 0.6978 0.9241 2.222
##
## Concordance= 0.653 (se = 0.033 )
## Likelihood ratio test= 49.34 on 3 df, p=1e-10
## Wald test = 232 on 3 df, p=<2e-16
## Score (logrank) test = 65.82 on 3 df, p=3e-14, Robust = 3.56 p=0.3
##
## (Note: the likelihood ratio and score tests assume independence of
## observations within a cluster, the Wald and robust score tests do not).cox.zph(cox_10_adj)## chisq df p
## strep.like.IMI.cow 0.5407 1 0.46
## calving.sau 0.0219 1 0.88
## calving.nas 0.2304 1 0.63
## GLOBAL 0.6684 3 0.88# Enterococcus
cox_11 <- coxph(survival~ enterococcus.spp.IMI.cow+cluster(Site), data=df_survival)
summary(cox_11)## Call:
## coxph(formula = survival ~ enterococcus.spp.IMI.cow, data = df_survival,
## cluster = Site)
##
## n= 333, number of events= 178
##
## coef exp(coef) se(coef) robust se z Pr(>|z|)
## enterococcus.spp.IMI.cow1 -0.2178 0.8043 0.2550 0.3109 -0.701 0.484
##
## exp(coef) exp(-coef) lower .95 upper .95
## enterococcus.spp.IMI.cow1 0.8043 1.243 0.4373 1.479
##
## Concordance= 0.509 (se = 0.011 )
## Likelihood ratio test= 0.77 on 1 df, p=0.4
## Wald test = 0.49 on 1 df, p=0.5
## Score (logrank) test = 0.73 on 1 df, p=0.4, Robust = 0.72 p=0.4
##
## (Note: the likelihood ratio and score tests assume independence of
## observations within a cluster, the Wald and robust score tests do not).cox.zph(cox_11)## chisq df p
## enterococcus.spp.IMI.cow 0.834 1 0.36
## GLOBAL 0.834 1 0.36cox_11_adj <- coxph(survival~ enterococcus.spp.IMI.cow+calving.sau+calving.nas+cluster(Site), data=df_survival)
summary(cox_11_adj)## Call:
## coxph(formula = survival ~ enterococcus.spp.IMI.cow + calving.sau +
## calving.nas, data = df_survival, cluster = Site)
##
## n= 333, number of events= 178
##
## coef exp(coef) se(coef) robust se z Pr(>|z|)
## enterococcus.spp.IMI.cow1 0.02545 1.02578 0.26057 0.31903 0.080 0.936
## calving.sau1 1.26437 3.54087 0.17178 0.14458 8.745 <2e-16
## calving.nas1 0.35433 1.42523 0.16116 0.23349 1.518 0.129
##
## enterococcus.spp.IMI.cow1
## calving.sau1 ***
## calving.nas1
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## exp(coef) exp(-coef) lower .95 upper .95
## enterococcus.spp.IMI.cow1 1.026 0.9749 0.5489 1.917
## calving.sau1 3.541 0.2824 2.6671 4.701
## calving.nas1 1.425 0.7016 0.9019 2.252
##
## Concordance= 0.653 (se = 0.034 )
## Likelihood ratio test= 49.22 on 3 df, p=1e-10
## Wald test = 270.9 on 3 df, p=<2e-16
## Score (logrank) test = 65.66 on 3 df, p=4e-14, Robust = 3.99 p=0.3
##
## (Note: the likelihood ratio and score tests assume independence of
## observations within a cluster, the Wald and robust score tests do not).cox.zph(cox_11_adj)## chisq df p
## enterococcus.spp.IMI.cow 1.0637 1 0.30
## calving.sau 0.0203 1 0.89
## calving.nas 0.2101 1 0.65
## GLOBAL 1.1460 3 0.77# Gram negative microorganisms
cox_12 <- coxph(survival~ gram.neg.IMI.cow+cluster(Site), data=df_survival)
summary(cox_12)## Call:
## coxph(formula = survival ~ gram.neg.IMI.cow, data = df_survival,
## cluster = Site)
##
## n= 333, number of events= 178
##
## coef exp(coef) se(coef) robust se z Pr(>|z|)
## gram.neg.IMI.cow1 -0.01514 0.98497 0.26209 0.25229 -0.06 0.952
##
## exp(coef) exp(-coef) lower .95 upper .95
## gram.neg.IMI.cow1 0.985 1.015 0.6007 1.615
##
## Concordance= 0.498 (se = 0.014 )
## Likelihood ratio test= 0 on 1 df, p=1
## Wald test = 0 on 1 df, p=1
## Score (logrank) test = 0 on 1 df, p=1, Robust = 0 p=1
##
## (Note: the likelihood ratio and score tests assume independence of
## observations within a cluster, the Wald and robust score tests do not).cox.zph(cox_12)## chisq df p
## gram.neg.IMI.cow 0.405 1 0.52
## GLOBAL 0.405 1 0.52cox_12_adj <- coxph(survival~ gram.neg.IMI.cow+calving.sau.sslo+calving.nas+cluster(Site), data=df_survival)
summary(cox_12_adj)## Call:
## coxph(formula = survival ~ gram.neg.IMI.cow + calving.sau.sslo +
## calving.nas, data = df_survival, cluster = Site)
##
## n= 333, number of events= 178
##
## coef exp(coef) se(coef) robust se z Pr(>|z|)
## gram.neg.IMI.cow1 0.1323 1.1414 0.2644 0.1948 0.679 0.4970
## calving.sau.sslo1 1.1758 3.2408 0.1527 0.1125 10.448 <2e-16 ***
## calving.nas1 0.4199 1.5218 0.1606 0.2385 1.760 0.0784 .
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## exp(coef) exp(-coef) lower .95 upper .95
## gram.neg.IMI.cow1 1.141 0.8761 0.7792 1.672
## calving.sau.sslo1 3.241 0.3086 2.5993 4.041
## calving.nas1 1.522 0.6571 0.9535 2.429
##
## Concordance= 0.691 (se = 0.035 )
## Likelihood ratio test= 60.2 on 3 df, p=5e-13
## Wald test = 187.5 on 3 df, p=<2e-16
## Score (logrank) test = 69.07 on 3 df, p=7e-15, Robust = 3.51 p=0.3
##
## (Note: the likelihood ratio and score tests assume independence of
## observations within a cluster, the Wald and robust score tests do not).cox.zph(cox_12_adj)## chisq df p
## gram.neg.IMI.cow 0.219 1 0.640
## calving.sau.sslo 4.310 1 0.038
## calving.nas 0.493 1 0.483
## GLOBAL 4.926 3 0.177# Others
cox_13 <- coxph(survival~ other.IMI.cow+cluster(Site), data=df_survival)
summary(cox_13)## Call:
## coxph(formula = survival ~ other.IMI.cow, data = df_survival,
## cluster = Site)
##
## n= 333, number of events= 178
##
## coef exp(coef) se(coef) robust se z Pr(>|z|)
## other.IMI.cow1 -0.6839 0.5046 0.1693 0.2998 -2.281 0.0225 *
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## exp(coef) exp(-coef) lower .95 upper .95
## other.IMI.cow1 0.5046 1.982 0.2804 0.9082
##
## Concordance= 0.586 (se = 0.043 )
## Likelihood ratio test= 17.84 on 1 df, p=2e-05
## Wald test = 5.2 on 1 df, p=0.02
## Score (logrank) test = 16.95 on 1 df, p=4e-05, Robust = 2.21 p=0.1
##
## (Note: the likelihood ratio and score tests assume independence of
## observations within a cluster, the Wald and robust score tests do not).cox.zph(cox_13)## chisq df p
## other.IMI.cow 0.0262 1 0.87
## GLOBAL 0.0262 1 0.87cox_13_adj <- coxph(survival~ other.IMI.cow+calving.sau.sslo+calving.nas+cluster(Site), data=df_survival)
summary(cox_13_adj)## Call:
## coxph(formula = survival ~ other.IMI.cow + calving.sau.sslo +
## calving.nas, data = df_survival, cluster = Site)
##
## n= 333, number of events= 178
##
## coef exp(coef) se(coef) robust se z Pr(>|z|)
## other.IMI.cow1 -0.58876 0.55501 0.17249 0.23388 -2.517 0.0118 *
## calving.sau.sslo1 1.14702 3.14880 0.15301 0.07535 15.223 <2e-16 ***
## calving.nas1 0.29637 1.34497 0.16231 0.14136 2.097 0.0360 *
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## exp(coef) exp(-coef) lower .95 upper .95
## other.IMI.cow1 0.555 1.8018 0.3509 0.8778
## calving.sau.sslo1 3.149 0.3176 2.7165 3.6499
## calving.nas1 1.345 0.7435 1.0195 1.7744
##
## Concordance= 0.717 (se = 0.035 )
## Likelihood ratio test= 72.48 on 3 df, p=1e-15
## Wald test = 435.4 on 3 df, p=<2e-16
## Score (logrank) test = 79.97 on 3 df, p=<2e-16, Robust = 3.62 p=0.3
##
## (Note: the likelihood ratio and score tests assume independence of
## observations within a cluster, the Wald and robust score tests do not).cox.zph(cox_13_adj)## chisq df p
## other.IMI.cow 0.155 1 0.694
## calving.sau.sslo 3.957 1 0.047
## calving.nas 0.410 1 0.522
## GLOBAL 4.271 3 0.234# Bacillus
cox_14 <- coxph(survival~ bacillus.IMI.cow+cluster(Site), data=df_survival)
summary(cox_14)## Call:
## coxph(formula = survival ~ bacillus.IMI.cow, data = df_survival,
## cluster = Site)
##
## n= 333, number of events= 178
##
## coef exp(coef) se(coef) robust se z Pr(>|z|)
## bacillus.IMI.cow1 -0.9696 0.3792 0.2199 0.2535 -3.826 0.00013 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## exp(coef) exp(-coef) lower .95 upper .95
## bacillus.IMI.cow1 0.3792 2.637 0.2308 0.6232
##
## Concordance= 0.591 (se = 0.038 )
## Likelihood ratio test= 24.31 on 1 df, p=8e-07
## Wald test = 14.64 on 1 df, p=1e-04
## Score (logrank) test = 21 on 1 df, p=5e-06, Robust = 2.33 p=0.1
##
## (Note: the likelihood ratio and score tests assume independence of
## observations within a cluster, the Wald and robust score tests do not).cox.zph(cox_14)## chisq df p
## bacillus.IMI.cow 0.71 1 0.4
## GLOBAL 0.71 1 0.4cox_14_adj <- coxph(survival~ bacillus.IMI.cow+calving.sau.sslo+calving.nas+cluster(Site), data=df_survival)
summary(cox_14_adj)## Call:
## coxph(formula = survival ~ bacillus.IMI.cow + calving.sau.sslo +
## calving.nas, data = df_survival, cluster = Site)
##
## n= 333, number of events= 178
##
## coef exp(coef) se(coef) robust se z Pr(>|z|)
## bacillus.IMI.cow1 -0.79221 0.45284 0.22241 0.21139 -3.748 0.000179 ***
## calving.sau.sslo1 1.09755 2.99683 0.15347 0.07407 14.818 < 2e-16 ***
## calving.nas1 0.32994 1.39088 0.16052 0.16713 1.974 0.048361 *
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## exp(coef) exp(-coef) lower .95 upper .95
## bacillus.IMI.cow1 0.4528 2.2083 0.2992 0.6853
## calving.sau.sslo1 2.9968 0.3337 2.5919 3.4650
## calving.nas1 1.3909 0.7190 1.0024 1.9300
##
## Concordance= 0.72 (se = 0.033 )
## Likelihood ratio test= 75.07 on 3 df, p=3e-16
## Wald test = 289.4 on 3 df, p=<2e-16
## Score (logrank) test = 80.63 on 3 df, p=<2e-16, Robust = 3.67 p=0.3
##
## (Note: the likelihood ratio and score tests assume independence of
## observations within a cluster, the Wald and robust score tests do not).cox.zph(cox_14_adj)## chisq df p
## bacillus.IMI.cow 0.417 1 0.52
## calving.sau.sslo 4.232 1 0.04
## calving.nas 0.541 1 0.46
## GLOBAL 4.778 3 0.19# Coryne
cox_15 <- coxph(survival~ coryne.IMI.cow+cluster(Site), data=df_survival)
summary(cox_15)## Call:
## coxph(formula = survival ~ coryne.IMI.cow, data = df_survival,
## cluster = Site)
##
## n= 333, number of events= 178
##
## coef exp(coef) se(coef) robust se z Pr(>|z|)
## coryne.IMI.cow1 -0.4069 0.6657 0.2280 0.3529 -1.153 0.249
##
## exp(coef) exp(-coef) lower .95 upper .95
## coryne.IMI.cow1 0.6657 1.502 0.3333 1.33
##
## Concordance= 0.524 (se = 0.033 )
## Likelihood ratio test= 3.52 on 1 df, p=0.06
## Wald test = 1.33 on 1 df, p=0.2
## Score (logrank) test = 3.23 on 1 df, p=0.07, Robust = 0.76 p=0.4
##
## (Note: the likelihood ratio and score tests assume independence of
## observations within a cluster, the Wald and robust score tests do not).cox.zph(cox_15)## chisq df p
## coryne.IMI.cow 2.04 1 0.15
## GLOBAL 2.04 1 0.15cox_15_adj <- coxph(survival~ coryne.IMI.cow+calving.sau.sslo+calving.nas+cluster(Site), data=df_survival)
summary(cox_15_adj)## Call:
## coxph(formula = survival ~ coryne.IMI.cow + calving.sau.sslo +
## calving.nas, data = df_survival, cluster = Site)
##
## n= 333, number of events= 178
##
## coef exp(coef) se(coef) robust se z Pr(>|z|)
## coryne.IMI.cow1 -0.3658 0.6936 0.2346 0.3115 -1.175 0.2402
## calving.sau.sslo1 1.1782 3.2486 0.1526 0.1098 10.727 <2e-16 ***
## calving.nas1 0.3442 1.4109 0.1635 0.1742 1.976 0.0481 *
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## exp(coef) exp(-coef) lower .95 upper .95
## coryne.IMI.cow1 0.6936 1.4417 0.3767 1.277
## calving.sau.sslo1 3.2486 0.3078 2.6194 4.029
## calving.nas1 1.4109 0.7088 1.0029 1.985
##
## Concordance= 0.69 (se = 0.04 )
## Likelihood ratio test= 62.59 on 3 df, p=2e-13
## Wald test = 322.3 on 3 df, p=<2e-16
## Score (logrank) test = 71.17 on 3 df, p=2e-15, Robust = 3.61 p=0.3
##
## (Note: the likelihood ratio and score tests assume independence of
## observations within a cluster, the Wald and robust score tests do not).cox.zph(cox_15_adj)## chisq df p
## coryne.IMI.cow 3.075 1 0.080
## calving.sau.sslo 3.992 1 0.046
## calving.nas 0.381 1 0.537
## GLOBAL 6.724 3 0.081fp_1 <- read.csv(file = "./data/forestplot/forestplot_hr_imi_scc.csv", header=TRUE)
fp_1<-fp_1 %>% rename("Crude risk exposed"=Crude.risk.exposed,"Crude risk unexposed"=Crude.risk.unexposed,"Taxonomic group"=Taxonomic.group)
fp_1$se <- (log(fp_1$hi) - log(fp_1$est))/1.96
fp_1$` ` <- paste(rep(" ", 15), collapse = " ")
fp_1$`HR (95% CI)` <- ifelse(is.na(fp_1$se), "",
sprintf("%.2f (%.2f to %.2f)",
fp_1$est, fp_1$low, fp_1$hi))
tm <- forest_theme(base_size = 9,xaxis_cex=0.8)
p <- forest(fp_1[,c(1:3, 8:9)],
est = fp_1$est,
lower = fp_1$low,
upper = fp_1$hi,
ci_column = 4,
ref_line = 1,
arrow_lab = c("Lower hazards high SCC", "Higher hazards high SCC"),
ticks_at = c(0.5, 1, 2,3,4),theme=tm)
plot(p)
ggsave(plot = p,"./figures_article_2/forestplot.png",width = 20, height = 12,dpi=300, units = "cm") 2.2 CA: Milk production
# ---- Staph aureus ----
model_1 <- lme(fixed =milk_kg~ staph.aureus.IMI.cow*postpartum_sample,random=~1|Site/CowId,correlation= corAR1(form = ~ps_sample),data=df.per.cow_dhi,na.action = na.omit)
emmeans(model_1,revpairwise~staph.aureus.IMI.cow|postpartum_sample)## $emmeans
## postpartum_sample = 1:
## staph.aureus.IMI.cow emmean SE df lower.CL upper.CL
## 0 23.5 3.21 3 13.3 33.7
## 1 23.7 3.28 3 13.3 34.2
##
## postpartum_sample = 2:
## staph.aureus.IMI.cow emmean SE df lower.CL upper.CL
## 0 25.6 3.21 3 15.4 35.9
## 1 25.4 3.28 3 15.0 35.9
##
## postpartum_sample = 3:
## staph.aureus.IMI.cow emmean SE df lower.CL upper.CL
## 0 24.8 3.21 3 14.5 35.0
## 1 24.7 3.29 3 14.2 35.1
##
## postpartum_sample = 4:
## staph.aureus.IMI.cow emmean SE df lower.CL upper.CL
## 0 23.7 3.21 3 13.5 33.9
## 1 23.8 3.29 3 13.3 34.3
##
## postpartum_sample = 5:
## staph.aureus.IMI.cow emmean SE df lower.CL upper.CL
## 0 23.1 3.21 3 12.9 33.3
## 1 23.3 3.29 3 12.8 33.7
##
## postpartum_sample = 6:
## staph.aureus.IMI.cow emmean SE df lower.CL upper.CL
## 0 20.9 3.21 3 10.7 31.1
## 1 21.4 3.30 3 10.9 31.8
##
## Degrees-of-freedom method: containment
## Confidence level used: 0.95
##
## $contrasts
## postpartum_sample = 1:
## contrast estimate SE df t.ratio
## staph.aureus.IMI.cow1 - staph.aureus.IMI.cow0 0.2491 0.946 328 0.263
## p.value
## 0.7924
##
## postpartum_sample = 2:
## contrast estimate SE df t.ratio
## staph.aureus.IMI.cow1 - staph.aureus.IMI.cow0 -0.2048 0.931 328 -0.220
## p.value
## 0.8260
##
## postpartum_sample = 3:
## contrast estimate SE df t.ratio
## staph.aureus.IMI.cow1 - staph.aureus.IMI.cow0 -0.0587 0.961 328 -0.061
## p.value
## 0.9513
##
## postpartum_sample = 4:
## contrast estimate SE df t.ratio
## staph.aureus.IMI.cow1 - staph.aureus.IMI.cow0 0.1328 1.005 328 0.132
## p.value
## 0.8950
##
## postpartum_sample = 5:
## contrast estimate SE df t.ratio
## staph.aureus.IMI.cow1 - staph.aureus.IMI.cow0 0.1804 0.989 328 0.182
## p.value
## 0.8554
##
## postpartum_sample = 6:
## contrast estimate SE df t.ratio
## staph.aureus.IMI.cow1 - staph.aureus.IMI.cow0 0.4445 1.004 328 0.443
## p.value
## 0.6582
##
## Degrees-of-freedom method: containmentmodel_1 <- lme(fixed =milk_kg~ staph.aureus.IMI.cow*postpartum_sample+calving.nas+calving.sslo,random=~1|Site/CowId,correlation= corAR1(form = ~ps_sample),data=df.per.cow_dhi,na.action = na.omit)
emmeans_1 <- emmeans(model_1,revpairwise~staph.aureus.IMI.cow|postpartum_sample)
emmeans_1 ## $emmeans
## postpartum_sample = 1:
## staph.aureus.IMI.cow emmean SE df lower.CL upper.CL
## 0 22.8 3.12 3 12.9 32.8
## 1 23.0 3.21 3 12.8 33.2
##
## postpartum_sample = 2:
## staph.aureus.IMI.cow emmean SE df lower.CL upper.CL
## 0 25.0 3.12 3 15.0 34.9
## 1 24.7 3.21 3 14.5 34.9
##
## postpartum_sample = 3:
## staph.aureus.IMI.cow emmean SE df lower.CL upper.CL
## 0 24.1 3.13 3 14.2 34.0
## 1 24.0 3.21 3 13.7 34.2
##
## postpartum_sample = 4:
## staph.aureus.IMI.cow emmean SE df lower.CL upper.CL
## 0 23.0 3.13 3 13.1 33.0
## 1 23.1 3.22 3 12.8 33.3
##
## postpartum_sample = 5:
## staph.aureus.IMI.cow emmean SE df lower.CL upper.CL
## 0 22.4 3.13 3 12.5 32.4
## 1 22.5 3.22 3 12.3 32.8
##
## postpartum_sample = 6:
## staph.aureus.IMI.cow emmean SE df lower.CL upper.CL
## 0 20.3 3.13 3 10.3 30.2
## 1 20.6 3.22 3 10.4 30.9
##
## Results are averaged over the levels of: calving.nas, calving.sslo
## Degrees-of-freedom method: containment
## Confidence level used: 0.95
##
## $contrasts
## postpartum_sample = 1:
## contrast estimate SE df t.ratio
## staph.aureus.IMI.cow1 - staph.aureus.IMI.cow0 0.1431 0.951 326 0.151
## p.value
## 0.8805
##
## postpartum_sample = 2:
## contrast estimate SE df t.ratio
## staph.aureus.IMI.cow1 - staph.aureus.IMI.cow0 -0.3027 0.936 326 -0.323
## p.value
## 0.7467
##
## postpartum_sample = 3:
## contrast estimate SE df t.ratio
## staph.aureus.IMI.cow1 - staph.aureus.IMI.cow0 -0.1473 0.966 326 -0.152
## p.value
## 0.8790
##
## postpartum_sample = 4:
## contrast estimate SE df t.ratio
## staph.aureus.IMI.cow1 - staph.aureus.IMI.cow0 0.0506 1.011 326 0.050
## p.value
## 0.9601
##
## postpartum_sample = 5:
## contrast estimate SE df t.ratio
## staph.aureus.IMI.cow1 - staph.aureus.IMI.cow0 0.1062 0.995 326 0.107
## p.value
## 0.9150
##
## postpartum_sample = 6:
## contrast estimate SE df t.ratio
## staph.aureus.IMI.cow1 - staph.aureus.IMI.cow0 0.3707 1.009 326 0.367
## p.value
## 0.7136
##
## Results are averaged over the levels of: calving.nas, calving.sslo
## Degrees-of-freedom method: containmentconfint(emmeans_1)## $emmeans
## postpartum_sample = 1:
## staph.aureus.IMI.cow emmean SE df lower.CL upper.CL
## 0 22.8 3.12 3 12.9 32.8
## 1 23.0 3.21 3 12.8 33.2
##
## postpartum_sample = 2:
## staph.aureus.IMI.cow emmean SE df lower.CL upper.CL
## 0 25.0 3.12 3 15.0 34.9
## 1 24.7 3.21 3 14.5 34.9
##
## postpartum_sample = 3:
## staph.aureus.IMI.cow emmean SE df lower.CL upper.CL
## 0 24.1 3.13 3 14.2 34.0
## 1 24.0 3.21 3 13.7 34.2
##
## postpartum_sample = 4:
## staph.aureus.IMI.cow emmean SE df lower.CL upper.CL
## 0 23.0 3.13 3 13.1 33.0
## 1 23.1 3.22 3 12.8 33.3
##
## postpartum_sample = 5:
## staph.aureus.IMI.cow emmean SE df lower.CL upper.CL
## 0 22.4 3.13 3 12.5 32.4
## 1 22.5 3.22 3 12.3 32.8
##
## postpartum_sample = 6:
## staph.aureus.IMI.cow emmean SE df lower.CL upper.CL
## 0 20.3 3.13 3 10.3 30.2
## 1 20.6 3.22 3 10.4 30.9
##
## Results are averaged over the levels of: calving.nas, calving.sslo
## Degrees-of-freedom method: containment
## Confidence level used: 0.95
##
## $contrasts
## postpartum_sample = 1:
## contrast estimate SE df lower.CL
## staph.aureus.IMI.cow1 - staph.aureus.IMI.cow0 0.1431 0.951 326 -1.73
## upper.CL
## 2.01
##
## postpartum_sample = 2:
## contrast estimate SE df lower.CL
## staph.aureus.IMI.cow1 - staph.aureus.IMI.cow0 -0.3027 0.936 326 -2.14
## upper.CL
## 1.54
##
## postpartum_sample = 3:
## contrast estimate SE df lower.CL
## staph.aureus.IMI.cow1 - staph.aureus.IMI.cow0 -0.1473 0.966 326 -2.05
## upper.CL
## 1.75
##
## postpartum_sample = 4:
## contrast estimate SE df lower.CL
## staph.aureus.IMI.cow1 - staph.aureus.IMI.cow0 0.0506 1.011 326 -1.94
## upper.CL
## 2.04
##
## postpartum_sample = 5:
## contrast estimate SE df lower.CL
## staph.aureus.IMI.cow1 - staph.aureus.IMI.cow0 0.1062 0.995 326 -1.85
## upper.CL
## 2.06
##
## postpartum_sample = 6:
## contrast estimate SE df lower.CL
## staph.aureus.IMI.cow1 - staph.aureus.IMI.cow0 0.3707 1.009 326 -1.61
## upper.CL
## 2.36
##
## Results are averaged over the levels of: calving.nas, calving.sslo
## Degrees-of-freedom method: containment
## Confidence level used: 0.95adj_means_1 <- as.data.frame(emmeans_1$emmeans)
adj_means_1## postpartum_sample = 1:
## staph.aureus.IMI.cow emmean SE df lower.CL upper.CL
## 0 22.84164 3.123483 3 12.90132 32.78196
## 1 22.98471 3.210114 3 12.76869 33.20073
##
## postpartum_sample = 2:
## staph.aureus.IMI.cow emmean SE df lower.CL upper.CL
## 0 24.98845 3.124083 3 15.04623 34.93068
## 1 24.68579 3.206451 3 14.48143 34.89015
##
## postpartum_sample = 3:
## staph.aureus.IMI.cow emmean SE df lower.CL upper.CL
## 0 24.09988 3.125252 3 14.15393 34.04582
## 1 23.95261 3.213118 3 13.72703 34.17818
##
## postpartum_sample = 4:
## staph.aureus.IMI.cow emmean SE df lower.CL upper.CL
## 0 23.03822 3.131872 3 13.07121 33.00524
## 1 23.08880 3.222302 3 12.83400 33.34360
##
## postpartum_sample = 5:
## staph.aureus.IMI.cow emmean SE df lower.CL upper.CL
## 0 22.43902 3.128446 3 12.48291 32.39513
## 1 22.54527 3.219222 3 12.30026 32.79027
##
## postpartum_sample = 6:
## staph.aureus.IMI.cow emmean SE df lower.CL upper.CL
## 0 20.25657 3.129886 3 10.29588 30.21726
## 1 20.62728 3.223071 3 10.37003 30.88453
##
## Results are averaged over the levels of: calving.nas, calving.sslo
## Degrees-of-freedom method: containment
## Confidence level used: 0.95contrasts_1 <- as.data.frame(emmeans_1$contrasts) %>% select (postpartum_sample,p.value)
adj_means_1 <- adj_means_1 %>% left_join(contrasts_1, by="postpartum_sample")
adj_means_1 <- adj_means_1 %>% group_by(postpartum_sample) %>% mutate(max_milk = max(upper.CL)) %>% ungroup()
adj_means_1 <- adj_means_1 %>% mutate(p.value = ifelse(max_milk == upper.CL, p.value, 1))
milk_1 <- ggplot(adj_means_1, aes(x=postpartum_sample, y=emmean, group=staph.aureus.IMI.cow, color=staph.aureus.IMI.cow)) +
geom_line() +
geom_point()+
geom_errorbar(aes(ymin=emmean-SE, ymax=emmean+SE), width=.2,
position=position_dodge(0.05)) +
ylim(0,32) +
labs(x="Monthly postpartum test",y="Milk production (Kg/day)",title="Staphylococcus aureus", color="IMI") +
theme_classic() +
theme(plot.title=element_text(face="italic")) +
scale_color_viridis(discrete = TRUE) +
scale_color_manual(labels = c("No", "Yes"),
values = c("#009999", "#990000")) +
geom_text(aes(label = ifelse(p.value < 0.05, "*", "")),
vjust = -3, size = 5, color = "black")## Scale for colour is already present.
## Adding another scale for colour, which will replace the existing scale.model_1_adj_overall <- lme(fixed =milk_kg~ staph.aureus.IMI.cow+calving.nas+calving.sslo,random=~1|Site/CowId,correlation= corAR1(form = ~ps_sample),data=df.per.cow_dhi,na.action = na.omit)
Confint(model_1_adj_overall)## Estimate 2.5 % 97.5 %
## (Intercept) 24.17762289 18.107459 30.2477868
## staph.aureus.IMI.cow1 0.02637572 -1.288632 1.3413838
## calving.nas1 -0.38547607 -1.600980 0.8300278
## calving.sslo1 -1.99431519 -3.327743 -0.6608876 # annotate("text", x=5, y=30, label= "Estimate (95%CI):",size=4)
# ---- NAS ----
model_2 <- lme(fixed =milk_kg~ nas.IMI.cow*postpartum_sample,random=~1|Site/CowId,correlation= corAR1(form = ~ps_sample),data=df.per.cow_dhi,na.action = na.omit)
emmeans(model_2,revpairwise~nas.IMI.cow|postpartum_sample)## $emmeans
## postpartum_sample = 1:
## nas.IMI.cow emmean SE df lower.CL upper.CL
## 0 23.0 3.22 3 12.72 33.2
## 1 24.2 3.23 3 13.89 34.4
##
## postpartum_sample = 2:
## nas.IMI.cow emmean SE df lower.CL upper.CL
## 0 25.9 3.22 3 15.63 36.1
## 1 25.3 3.23 3 15.04 35.6
##
## postpartum_sample = 3:
## nas.IMI.cow emmean SE df lower.CL upper.CL
## 0 25.3 3.23 3 15.01 35.5
## 1 24.2 3.23 3 13.94 34.5
##
## postpartum_sample = 4:
## nas.IMI.cow emmean SE df lower.CL upper.CL
## 0 23.7 3.24 3 13.35 34.0
## 1 23.8 3.23 3 13.48 34.1
##
## postpartum_sample = 5:
## nas.IMI.cow emmean SE df lower.CL upper.CL
## 0 23.4 3.23 3 13.09 33.7
## 1 22.9 3.23 3 12.64 33.2
##
## postpartum_sample = 6:
## nas.IMI.cow emmean SE df lower.CL upper.CL
## 0 20.0 3.23 3 9.76 30.3
## 1 21.9 3.23 3 11.64 32.2
##
## Degrees-of-freedom method: containment
## Confidence level used: 0.95
##
## $contrasts
## postpartum_sample = 1:
## contrast estimate SE df t.ratio p.value
## nas.IMI.cow1 - nas.IMI.cow0 1.188 0.775 328 1.532 0.1264
##
## postpartum_sample = 2:
## contrast estimate SE df t.ratio p.value
## nas.IMI.cow1 - nas.IMI.cow0 -0.585 0.776 328 -0.754 0.4515
##
## postpartum_sample = 3:
## contrast estimate SE df t.ratio p.value
## nas.IMI.cow1 - nas.IMI.cow0 -1.066 0.797 328 -1.337 0.1822
##
## postpartum_sample = 4:
## contrast estimate SE df t.ratio p.value
## nas.IMI.cow1 - nas.IMI.cow0 0.116 0.853 328 0.136 0.8916
##
## postpartum_sample = 5:
## contrast estimate SE df t.ratio p.value
## nas.IMI.cow1 - nas.IMI.cow0 -0.445 0.827 328 -0.538 0.5912
##
## postpartum_sample = 6:
## contrast estimate SE df t.ratio p.value
## nas.IMI.cow1 - nas.IMI.cow0 1.887 0.841 328 2.245 0.0255
##
## Degrees-of-freedom method: containmentmodel_2 <- lme(fixed =milk_kg~ nas.IMI.cow*postpartum_sample+calving.sau+calving.sslo,random=~1|Site/CowId,correlation= corAR1(form = ~ps_sample),data=df.per.cow_dhi,na.action = na.omit)
emmeans_2 <- emmeans(model_2,revpairwise~nas.IMI.cow|postpartum_sample)
emmeans_2## $emmeans
## postpartum_sample = 1:
## nas.IMI.cow emmean SE df lower.CL upper.CL
## 0 22.5 3.16 3 12.40 32.5
## 1 23.4 3.17 3 13.35 33.5
##
## postpartum_sample = 2:
## nas.IMI.cow emmean SE df lower.CL upper.CL
## 0 25.4 3.16 3 15.32 35.4
## 1 24.6 3.17 3 14.48 34.7
##
## postpartum_sample = 3:
## nas.IMI.cow emmean SE df lower.CL upper.CL
## 0 24.8 3.16 3 14.71 34.8
## 1 23.5 3.17 3 13.39 33.6
##
## postpartum_sample = 4:
## nas.IMI.cow emmean SE df lower.CL upper.CL
## 0 23.1 3.17 3 13.04 33.2
## 1 23.0 3.18 3 12.92 33.2
##
## postpartum_sample = 5:
## nas.IMI.cow emmean SE df lower.CL upper.CL
## 0 22.9 3.17 3 12.79 32.9
## 1 22.2 3.18 3 12.09 32.3
##
## postpartum_sample = 6:
## nas.IMI.cow emmean SE df lower.CL upper.CL
## 0 19.5 3.17 3 9.46 29.6
## 1 21.2 3.18 3 11.09 31.3
##
## Results are averaged over the levels of: calving.sau, calving.sslo
## Degrees-of-freedom method: containment
## Confidence level used: 0.95
##
## $contrasts
## postpartum_sample = 1:
## contrast estimate SE df t.ratio p.value
## nas.IMI.cow1 - nas.IMI.cow0 0.987 0.779 326 1.268 0.2056
##
## postpartum_sample = 2:
## contrast estimate SE df t.ratio p.value
## nas.IMI.cow1 - nas.IMI.cow0 -0.801 0.780 326 -1.026 0.3055
##
## postpartum_sample = 3:
## contrast estimate SE df t.ratio p.value
## nas.IMI.cow1 - nas.IMI.cow0 -1.284 0.801 326 -1.603 0.1099
##
## postpartum_sample = 4:
## contrast estimate SE df t.ratio p.value
## nas.IMI.cow1 - nas.IMI.cow0 -0.102 0.857 326 -0.119 0.9053
##
## postpartum_sample = 5:
## contrast estimate SE df t.ratio p.value
## nas.IMI.cow1 - nas.IMI.cow0 -0.669 0.832 326 -0.804 0.4217
##
## postpartum_sample = 6:
## contrast estimate SE df t.ratio p.value
## nas.IMI.cow1 - nas.IMI.cow0 1.663 0.845 326 1.968 0.0499
##
## Results are averaged over the levels of: calving.sau, calving.sslo
## Degrees-of-freedom method: containmentconfint(emmeans_2)## $emmeans
## postpartum_sample = 1:
## nas.IMI.cow emmean SE df lower.CL upper.CL
## 0 22.5 3.16 3 12.40 32.5
## 1 23.4 3.17 3 13.35 33.5
##
## postpartum_sample = 2:
## nas.IMI.cow emmean SE df lower.CL upper.CL
## 0 25.4 3.16 3 15.32 35.4
## 1 24.6 3.17 3 14.48 34.7
##
## postpartum_sample = 3:
## nas.IMI.cow emmean SE df lower.CL upper.CL
## 0 24.8 3.16 3 14.71 34.8
## 1 23.5 3.17 3 13.39 33.6
##
## postpartum_sample = 4:
## nas.IMI.cow emmean SE df lower.CL upper.CL
## 0 23.1 3.17 3 13.04 33.2
## 1 23.0 3.18 3 12.92 33.2
##
## postpartum_sample = 5:
## nas.IMI.cow emmean SE df lower.CL upper.CL
## 0 22.9 3.17 3 12.79 32.9
## 1 22.2 3.18 3 12.09 32.3
##
## postpartum_sample = 6:
## nas.IMI.cow emmean SE df lower.CL upper.CL
## 0 19.5 3.17 3 9.46 29.6
## 1 21.2 3.18 3 11.09 31.3
##
## Results are averaged over the levels of: calving.sau, calving.sslo
## Degrees-of-freedom method: containment
## Confidence level used: 0.95
##
## $contrasts
## postpartum_sample = 1:
## contrast estimate SE df lower.CL upper.CL
## nas.IMI.cow1 - nas.IMI.cow0 0.987 0.779 326 -0.544077 2.519
##
## postpartum_sample = 2:
## contrast estimate SE df lower.CL upper.CL
## nas.IMI.cow1 - nas.IMI.cow0 -0.801 0.780 326 -2.335868 0.734
##
## postpartum_sample = 3:
## contrast estimate SE df lower.CL upper.CL
## nas.IMI.cow1 - nas.IMI.cow0 -1.284 0.801 326 -2.859554 0.292
##
## postpartum_sample = 4:
## contrast estimate SE df lower.CL upper.CL
## nas.IMI.cow1 - nas.IMI.cow0 -0.102 0.857 326 -1.788968 1.585
##
## postpartum_sample = 5:
## contrast estimate SE df lower.CL upper.CL
## nas.IMI.cow1 - nas.IMI.cow0 -0.669 0.832 326 -2.305188 0.967
##
## postpartum_sample = 6:
## contrast estimate SE df lower.CL upper.CL
## nas.IMI.cow1 - nas.IMI.cow0 1.663 0.845 326 0.000863 3.326
##
## Results are averaged over the levels of: calving.sau, calving.sslo
## Degrees-of-freedom method: containment
## Confidence level used: 0.95adj_means_2 <- as.data.frame(emmeans_2$emmeans)
adj_means_2## postpartum_sample = 1:
## nas.IMI.cow emmean SE df lower.CL upper.CL
## 0 22.45393 3.157673 3 12.404810 32.50306
## 1 23.44143 3.170286 3 13.352166 33.53069
##
## postpartum_sample = 2:
## nas.IMI.cow emmean SE df lower.CL upper.CL
## 0 25.37221 3.158365 3 15.320888 35.42354
## 1 24.57136 3.170636 3 14.480982 34.66174
##
## postpartum_sample = 3:
## nas.IMI.cow emmean SE df lower.CL upper.CL
## 0 24.76812 3.161398 3 14.707146 34.82910
## 1 23.48424 3.172571 3 13.387706 33.58078
##
## postpartum_sample = 4:
## nas.IMI.cow emmean SE df lower.CL upper.CL
## 0 23.14060 3.173148 3 13.042224 33.23897
## 1 23.03846 3.177895 3 12.924979 33.15194
##
## postpartum_sample = 5:
## nas.IMI.cow emmean SE df lower.CL upper.CL
## 0 22.86545 3.166509 3 12.788207 32.94270
## 1 22.19638 3.176077 3 12.088685 32.30407
##
## postpartum_sample = 6:
## nas.IMI.cow emmean SE df lower.CL upper.CL
## 0 19.53763 3.167748 3 9.456442 29.61882
## 1 21.20086 3.178637 3 11.085016 31.31670
##
## Results are averaged over the levels of: calving.sau, calving.sslo
## Degrees-of-freedom method: containment
## Confidence level used: 0.95contrasts_2 <- as.data.frame(emmeans_2$contrasts) %>% select (postpartum_sample,p.value)
adj_means_2 <- adj_means_2 %>% left_join(contrasts_2, by="postpartum_sample")
adj_means_2 <- adj_means_2 %>% group_by(postpartum_sample) %>% mutate(max_milk = max(upper.CL)) %>% ungroup()
adj_means_2 <- adj_means_2 %>% mutate(p.value = ifelse(max_milk == upper.CL, p.value, 1))
milk_2 <- ggplot(adj_means_2, aes(x=postpartum_sample, y=emmean, group=nas.IMI.cow, color=nas.IMI.cow)) +
geom_line() +
geom_point()+
geom_errorbar(aes(ymin=emmean-SE, ymax=emmean+SE), width=.2,
position=position_dodge(0.05)) +
ylim(0,32) +
labs(x="Monthly postpartum test",y="Milk production (Kg/day)",title="NASM",color="IMI") +
theme_classic() +
scale_color_viridis(discrete = TRUE) +
scale_color_manual(labels = c("No", "Yes"),
values = c("#009999", "#990000")) +
geom_text(aes(label = ifelse(p.value < 0.05, "*", "")),
vjust = -3, size = 5, color = "black")## Scale for colour is already present.
## Adding another scale for colour, which will replace the existing scale.model_2_adj_overall <- lme(fixed =milk_kg~ nas.IMI.cow+calving.sau+calving.sslo,random=~1|Site/CowId,correlation= corAR1(form = ~ps_sample),data=df.per.cow_dhi,na.action = na.omit)
Confint(model_2_adj_overall)## Estimate 2.5 % 97.5 %
## (Intercept) 24.04828039 17.948076 30.1484847
## nas.IMI.cow1 0.06725118 -1.015129 1.1496314
## calving.sau1 -0.14861744 -1.565902 1.2686667
## calving.sslo1 -1.91491994 -3.241671 -0.5881688# ---- Staph chromogenes ----
model_3 <- lme(fixed =milk_kg~ sch.IMI.cow*postpartum_sample,random=~1|Site/CowId,correlation= corAR1(form = ~ps_sample),data=df.per.cow_dhi,na.action = na.omit)
emmeans(model_3,revpairwise~sch.IMI.cow|postpartum_sample)## $emmeans
## postpartum_sample = 1:
## sch.IMI.cow emmean SE df lower.CL upper.CL
## 0 23.2 3.21 3 13.0 33.4
## 1 24.2 3.24 3 13.9 34.5
##
## postpartum_sample = 2:
## sch.IMI.cow emmean SE df lower.CL upper.CL
## 0 25.7 3.21 3 15.4 35.9
## 1 25.5 3.24 3 15.2 35.8
##
## postpartum_sample = 3:
## sch.IMI.cow emmean SE df lower.CL upper.CL
## 0 25.3 3.21 3 15.1 35.6
## 1 23.7 3.24 3 13.4 34.0
##
## postpartum_sample = 4:
## sch.IMI.cow emmean SE df lower.CL upper.CL
## 0 23.7 3.22 3 13.4 33.9
## 1 23.7 3.25 3 13.4 34.1
##
## postpartum_sample = 5:
## sch.IMI.cow emmean SE df lower.CL upper.CL
## 0 23.6 3.22 3 13.4 33.8
## 1 22.4 3.24 3 12.0 32.7
##
## postpartum_sample = 6:
## sch.IMI.cow emmean SE df lower.CL upper.CL
## 0 20.7 3.22 3 10.4 30.9
## 1 21.6 3.25 3 11.2 31.9
##
## Degrees-of-freedom method: containment
## Confidence level used: 0.95
##
## $contrasts
## postpartum_sample = 1:
## contrast estimate SE df t.ratio p.value
## sch.IMI.cow1 - sch.IMI.cow0 0.9706 0.804 328 1.207 0.2281
##
## postpartum_sample = 2:
## contrast estimate SE df t.ratio p.value
## sch.IMI.cow1 - sch.IMI.cow0 -0.1638 0.800 328 -0.205 0.8379
##
## postpartum_sample = 3:
## contrast estimate SE df t.ratio p.value
## sch.IMI.cow1 - sch.IMI.cow0 -1.6556 0.817 328 -2.026 0.0436
##
## postpartum_sample = 4:
## contrast estimate SE df t.ratio p.value
## sch.IMI.cow1 - sch.IMI.cow0 0.0422 0.867 328 0.049 0.9612
##
## postpartum_sample = 5:
## contrast estimate SE df t.ratio p.value
## sch.IMI.cow1 - sch.IMI.cow0 -1.2375 0.846 328 -1.463 0.1444
##
## postpartum_sample = 6:
## contrast estimate SE df t.ratio p.value
## sch.IMI.cow1 - sch.IMI.cow0 0.8884 0.863 328 1.030 0.3038
##
## Degrees-of-freedom method: containmentmodel_3 <- lme(fixed =milk_kg~ sch.IMI.cow*postpartum_sample+calving.sau+calving.sslo,random=~1|Site/CowId,correlation= corAR1(form = ~ps_sample),data=df.per.cow_dhi,na.action = na.omit)
emmeans_3 <- emmeans(model_3,revpairwise~sch.IMI.cow|postpartum_sample)
emmeans_3## $emmeans
## postpartum_sample = 1:
## sch.IMI.cow emmean SE df lower.CL upper.CL
## 0 22.6 3.14 3 12.6 32.6
## 1 23.4 3.18 3 13.3 33.6
##
## postpartum_sample = 2:
## sch.IMI.cow emmean SE df lower.CL upper.CL
## 0 25.1 3.14 3 15.1 35.1
## 1 24.7 3.18 3 14.6 34.8
##
## postpartum_sample = 3:
## sch.IMI.cow emmean SE df lower.CL upper.CL
## 0 24.8 3.14 3 14.8 34.8
## 1 22.9 3.18 3 12.8 33.1
##
## postpartum_sample = 4:
## sch.IMI.cow emmean SE df lower.CL upper.CL
## 0 23.1 3.15 3 13.1 33.1
## 1 23.0 3.19 3 12.8 33.1
##
## postpartum_sample = 5:
## sch.IMI.cow emmean SE df lower.CL upper.CL
## 0 23.0 3.15 3 13.0 33.0
## 1 21.6 3.19 3 11.5 31.7
##
## postpartum_sample = 6:
## sch.IMI.cow emmean SE df lower.CL upper.CL
## 0 20.1 3.15 3 10.1 30.1
## 1 20.8 3.19 3 10.7 31.0
##
## Results are averaged over the levels of: calving.sau, calving.sslo
## Degrees-of-freedom method: containment
## Confidence level used: 0.95
##
## $contrasts
## postpartum_sample = 1:
## contrast estimate SE df t.ratio p.value
## sch.IMI.cow1 - sch.IMI.cow0 0.802 0.804 326 0.998 0.3191
##
## postpartum_sample = 2:
## contrast estimate SE df t.ratio p.value
## sch.IMI.cow1 - sch.IMI.cow0 -0.343 0.801 326 -0.428 0.6689
##
## postpartum_sample = 3:
## contrast estimate SE df t.ratio p.value
## sch.IMI.cow1 - sch.IMI.cow0 -1.842 0.818 326 -2.253 0.0249
##
## postpartum_sample = 4:
## contrast estimate SE df t.ratio p.value
## sch.IMI.cow1 - sch.IMI.cow0 -0.140 0.867 326 -0.162 0.8715
##
## postpartum_sample = 5:
## contrast estimate SE df t.ratio p.value
## sch.IMI.cow1 - sch.IMI.cow0 -1.423 0.846 326 -1.682 0.0936
##
## postpartum_sample = 6:
## contrast estimate SE df t.ratio p.value
## sch.IMI.cow1 - sch.IMI.cow0 0.697 0.863 326 0.808 0.4199
##
## Results are averaged over the levels of: calving.sau, calving.sslo
## Degrees-of-freedom method: containmentconfint(emmeans_3)## $emmeans
## postpartum_sample = 1:
## sch.IMI.cow emmean SE df lower.CL upper.CL
## 0 22.6 3.14 3 12.6 32.6
## 1 23.4 3.18 3 13.3 33.6
##
## postpartum_sample = 2:
## sch.IMI.cow emmean SE df lower.CL upper.CL
## 0 25.1 3.14 3 15.1 35.1
## 1 24.7 3.18 3 14.6 34.8
##
## postpartum_sample = 3:
## sch.IMI.cow emmean SE df lower.CL upper.CL
## 0 24.8 3.14 3 14.8 34.8
## 1 22.9 3.18 3 12.8 33.1
##
## postpartum_sample = 4:
## sch.IMI.cow emmean SE df lower.CL upper.CL
## 0 23.1 3.15 3 13.1 33.1
## 1 23.0 3.19 3 12.8 33.1
##
## postpartum_sample = 5:
## sch.IMI.cow emmean SE df lower.CL upper.CL
## 0 23.0 3.15 3 13.0 33.0
## 1 21.6 3.19 3 11.5 31.7
##
## postpartum_sample = 6:
## sch.IMI.cow emmean SE df lower.CL upper.CL
## 0 20.1 3.15 3 10.1 30.1
## 1 20.8 3.19 3 10.7 31.0
##
## Results are averaged over the levels of: calving.sau, calving.sslo
## Degrees-of-freedom method: containment
## Confidence level used: 0.95
##
## $contrasts
## postpartum_sample = 1:
## contrast estimate SE df lower.CL upper.CL
## sch.IMI.cow1 - sch.IMI.cow0 0.802 0.804 326 -0.779 2.383
##
## postpartum_sample = 2:
## contrast estimate SE df lower.CL upper.CL
## sch.IMI.cow1 - sch.IMI.cow0 -0.343 0.801 326 -1.918 1.233
##
## postpartum_sample = 3:
## contrast estimate SE df lower.CL upper.CL
## sch.IMI.cow1 - sch.IMI.cow0 -1.842 0.818 326 -3.450 -0.234
##
## postpartum_sample = 4:
## contrast estimate SE df lower.CL upper.CL
## sch.IMI.cow1 - sch.IMI.cow0 -0.140 0.867 326 -1.847 1.566
##
## postpartum_sample = 5:
## contrast estimate SE df lower.CL upper.CL
## sch.IMI.cow1 - sch.IMI.cow0 -1.423 0.846 326 -3.089 0.242
##
## postpartum_sample = 6:
## contrast estimate SE df lower.CL upper.CL
## sch.IMI.cow1 - sch.IMI.cow0 0.697 0.863 326 -1.001 2.396
##
## Results are averaged over the levels of: calving.sau, calving.sslo
## Degrees-of-freedom method: containment
## Confidence level used: 0.95adj_means_3 <- as.data.frame(emmeans_3$emmeans)
adj_means_3## postpartum_sample = 1:
## sch.IMI.cow emmean SE df lower.CL upper.CL
## 0 22.63204 3.141625 3 12.63398 32.63009
## 1 23.43412 3.179063 3 13.31692 33.55132
##
## postpartum_sample = 2:
## sch.IMI.cow emmean SE df lower.CL upper.CL
## 0 25.06914 3.142091 3 15.06961 35.06868
## 1 24.72630 3.178837 3 14.60983 34.84278
##
## postpartum_sample = 3:
## sch.IMI.cow emmean SE df lower.CL upper.CL
## 0 24.77413 3.144591 3 14.76664 34.78162
## 1 22.93216 3.180489 3 12.81043 33.05390
##
## postpartum_sample = 4:
## sch.IMI.cow emmean SE df lower.CL upper.CL
## 0 23.10860 3.152611 3 13.07558 33.14161
## 1 22.96818 3.187903 3 12.82285 33.11351
##
## postpartum_sample = 5:
## sch.IMI.cow emmean SE df lower.CL upper.CL
## 0 23.02810 3.148298 3 13.00881 33.04739
## 1 21.60471 3.185213 3 11.46794 31.74148
##
## postpartum_sample = 6:
## sch.IMI.cow emmean SE df lower.CL upper.CL
## 0 20.10806 3.149439 3 10.08513 30.13098
## 1 20.80530 3.188879 3 10.65686 30.95374
##
## Results are averaged over the levels of: calving.sau, calving.sslo
## Degrees-of-freedom method: containment
## Confidence level used: 0.95contrasts_3 <- as.data.frame(emmeans_3$contrasts) %>% select (postpartum_sample,p.value)
adj_means_3 <- adj_means_3 %>% left_join(contrasts_3, by="postpartum_sample")
adj_means_3 <- adj_means_3 %>% group_by(postpartum_sample) %>% mutate(max_milk = max(upper.CL)) %>% ungroup()
adj_means_3 <- adj_means_3 %>% mutate(p.value = ifelse(max_milk == upper.CL, p.value, 1))
milk_3 <- ggplot(adj_means_3, aes(x=postpartum_sample, y=emmean, group=sch.IMI.cow, color=sch.IMI.cow)) +
geom_line() +
geom_point()+
geom_errorbar(aes(ymin=emmean-SE, ymax=emmean+SE), width=.2,
position=position_dodge(0.05)) +
ylim(0,32) +
labs(x="Monthly postpartum test",y="Milk production (Kg/day)",title="Staphylococcus chromogenes",color="IMI") +
theme_classic() +
theme(plot.title=element_text(face="italic")) +
scale_color_viridis(discrete = TRUE) +
scale_color_manual(labels = c("No", "Yes"),
values = c("#009999", "#990000")) +
geom_text(aes(label = ifelse(p.value < 0.05, "*", "")),
vjust = -3, size = 5, color = "black")## Scale for colour is already present.
## Adding another scale for colour, which will replace the existing scale.model_3_adj_overall <- lme(fixed =milk_kg~ sch.IMI.cow+calving.sau+calving.sslo,random=~1|Site/CowId,correlation= corAR1(form = ~ps_sample),data=df.per.cow_dhi,na.action = na.omit)
Confint(model_3_adj_overall)## Estimate 2.5 % 97.5 %
## (Intercept) 24.2075407 18.132064 30.2830170
## sch.IMI.cow1 -0.3057573 -1.397148 0.7856335
## calving.sau1 -0.1927147 -1.602221 1.2167920
## calving.sslo1 -1.9648252 -3.286422 -0.6432286# ---- NAS non chromogenes -----
model_4 <- lme(fixed =milk_kg~ nas.non.chrom.IMI.cow*postpartum_sample,random=~1|Site/CowId,correlation= corAR1(form = ~ps_sample),data=df.per.cow_dhi,na.action = na.omit)
emmeans(model_4,revpairwise~nas.non.chrom.IMI.cow|postpartum_sample)## $emmeans
## postpartum_sample = 1:
## nas.non.chrom.IMI.cow emmean SE df lower.CL upper.CL
## 0 23.3 3.22 3 13.1 33.6
## 1 24.4 3.30 3 13.9 34.9
##
## postpartum_sample = 2:
## nas.non.chrom.IMI.cow emmean SE df lower.CL upper.CL
## 0 25.7 3.22 3 15.4 35.9
## 1 25.4 3.29 3 14.9 35.9
##
## postpartum_sample = 3:
## nas.non.chrom.IMI.cow emmean SE df lower.CL upper.CL
## 0 24.7 3.22 3 14.5 35.0
## 1 24.8 3.30 3 14.3 35.3
##
## postpartum_sample = 4:
## nas.non.chrom.IMI.cow emmean SE df lower.CL upper.CL
## 0 23.6 3.22 3 13.4 33.9
## 1 24.2 3.30 3 13.7 34.7
##
## postpartum_sample = 5:
## nas.non.chrom.IMI.cow emmean SE df lower.CL upper.CL
## 0 22.8 3.22 3 12.5 33.1
## 1 24.3 3.30 3 13.8 34.8
##
## postpartum_sample = 6:
## nas.non.chrom.IMI.cow emmean SE df lower.CL upper.CL
## 0 20.5 3.22 3 10.3 30.8
## 1 22.7 3.31 3 12.2 33.2
##
## Degrees-of-freedom method: containment
## Confidence level used: 0.95
##
## $contrasts
## postpartum_sample = 1:
## contrast estimate SE df t.ratio
## nas.non.chrom.IMI.cow1 - nas.non.chrom.IMI.cow0 1.117 0.962 328 1.161
## p.value
## 0.2465
##
## postpartum_sample = 2:
## contrast estimate SE df t.ratio
## nas.non.chrom.IMI.cow1 - nas.non.chrom.IMI.cow0 -0.252 0.939 328 -0.268
## p.value
## 0.7890
##
## postpartum_sample = 3:
## contrast estimate SE df t.ratio
## nas.non.chrom.IMI.cow1 - nas.non.chrom.IMI.cow0 0.113 0.956 328 0.118
## p.value
## 0.9060
##
## postpartum_sample = 4:
## contrast estimate SE df t.ratio
## nas.non.chrom.IMI.cow1 - nas.non.chrom.IMI.cow0 0.555 0.991 328 0.560
## p.value
## 0.5759
##
## postpartum_sample = 5:
## contrast estimate SE df t.ratio
## nas.non.chrom.IMI.cow1 - nas.non.chrom.IMI.cow0 1.494 0.981 328 1.523
## p.value
## 0.1288
##
## postpartum_sample = 6:
## contrast estimate SE df t.ratio
## nas.non.chrom.IMI.cow1 - nas.non.chrom.IMI.cow0 2.170 1.007 328 2.155
## p.value
## 0.0319
##
## Degrees-of-freedom method: containmentmodel_4 <- lme(fixed =milk_kg~ nas.non.chrom.IMI.cow*postpartum_sample+calving.sau+calving.sslo,random=~1|Site/CowId,correlation= corAR1(form = ~ps_sample),data=df.per.cow_dhi,na.action = na.omit)
emmeans_4 <- emmeans(model_4,revpairwise~nas.non.chrom.IMI.cow|postpartum_sample)
emmeans_4## $emmeans
## postpartum_sample = 1:
## nas.non.chrom.IMI.cow emmean SE df lower.CL upper.CL
## 0 22.8 3.16 3 12.73 32.8
## 1 23.8 3.25 3 13.47 34.2
##
## postpartum_sample = 2:
## nas.non.chrom.IMI.cow emmean SE df lower.CL upper.CL
## 0 25.1 3.16 3 15.07 35.2
## 1 24.8 3.25 3 14.46 35.1
##
## postpartum_sample = 3:
## nas.non.chrom.IMI.cow emmean SE df lower.CL upper.CL
## 0 24.2 3.16 3 14.13 34.2
## 1 24.2 3.25 3 13.88 34.6
##
## postpartum_sample = 4:
## nas.non.chrom.IMI.cow emmean SE df lower.CL upper.CL
## 0 23.1 3.17 3 13.00 33.1
## 1 23.5 3.25 3 13.19 33.9
##
## postpartum_sample = 5:
## nas.non.chrom.IMI.cow emmean SE df lower.CL upper.CL
## 0 22.3 3.16 3 12.20 32.3
## 1 23.7 3.25 3 13.31 34.0
##
## postpartum_sample = 6:
## nas.non.chrom.IMI.cow emmean SE df lower.CL upper.CL
## 0 20.0 3.16 3 9.93 30.1
## 1 22.1 3.26 3 11.70 32.5
##
## Results are averaged over the levels of: calving.sau, calving.sslo
## Degrees-of-freedom method: containment
## Confidence level used: 0.95
##
## $contrasts
## postpartum_sample = 1:
## contrast estimate SE df t.ratio
## nas.non.chrom.IMI.cow1 - nas.non.chrom.IMI.cow0 1.0435 0.964 326 1.082
## p.value
## 0.2800
##
## postpartum_sample = 2:
## contrast estimate SE df t.ratio
## nas.non.chrom.IMI.cow1 - nas.non.chrom.IMI.cow0 -0.3287 0.942 326 -0.349
## p.value
## 0.7273
##
## postpartum_sample = 3:
## contrast estimate SE df t.ratio
## nas.non.chrom.IMI.cow1 - nas.non.chrom.IMI.cow0 0.0358 0.959 326 0.037
## p.value
## 0.9702
##
## postpartum_sample = 4:
## contrast estimate SE df t.ratio
## nas.non.chrom.IMI.cow1 - nas.non.chrom.IMI.cow0 0.4758 0.994 326 0.479
## p.value
## 0.6325
##
## postpartum_sample = 5:
## contrast estimate SE df t.ratio
## nas.non.chrom.IMI.cow1 - nas.non.chrom.IMI.cow0 1.4090 0.984 326 1.432
## p.value
## 0.1531
##
## postpartum_sample = 6:
## contrast estimate SE df t.ratio
## nas.non.chrom.IMI.cow1 - nas.non.chrom.IMI.cow0 2.0768 1.009 326 2.057
## p.value
## 0.0404
##
## Results are averaged over the levels of: calving.sau, calving.sslo
## Degrees-of-freedom method: containmentconfint(emmeans_4)## $emmeans
## postpartum_sample = 1:
## nas.non.chrom.IMI.cow emmean SE df lower.CL upper.CL
## 0 22.8 3.16 3 12.73 32.8
## 1 23.8 3.25 3 13.47 34.2
##
## postpartum_sample = 2:
## nas.non.chrom.IMI.cow emmean SE df lower.CL upper.CL
## 0 25.1 3.16 3 15.07 35.2
## 1 24.8 3.25 3 14.46 35.1
##
## postpartum_sample = 3:
## nas.non.chrom.IMI.cow emmean SE df lower.CL upper.CL
## 0 24.2 3.16 3 14.13 34.2
## 1 24.2 3.25 3 13.88 34.6
##
## postpartum_sample = 4:
## nas.non.chrom.IMI.cow emmean SE df lower.CL upper.CL
## 0 23.1 3.17 3 13.00 33.1
## 1 23.5 3.25 3 13.19 33.9
##
## postpartum_sample = 5:
## nas.non.chrom.IMI.cow emmean SE df lower.CL upper.CL
## 0 22.3 3.16 3 12.20 32.3
## 1 23.7 3.25 3 13.31 34.0
##
## postpartum_sample = 6:
## nas.non.chrom.IMI.cow emmean SE df lower.CL upper.CL
## 0 20.0 3.16 3 9.93 30.1
## 1 22.1 3.26 3 11.70 32.5
##
## Results are averaged over the levels of: calving.sau, calving.sslo
## Degrees-of-freedom method: containment
## Confidence level used: 0.95
##
## $contrasts
## postpartum_sample = 1:
## contrast estimate SE df lower.CL
## nas.non.chrom.IMI.cow1 - nas.non.chrom.IMI.cow0 1.0435 0.964 326 -0.853
## upper.CL
## 2.94
##
## postpartum_sample = 2:
## contrast estimate SE df lower.CL
## nas.non.chrom.IMI.cow1 - nas.non.chrom.IMI.cow0 -0.3287 0.942 326 -2.181
## upper.CL
## 1.52
##
## postpartum_sample = 3:
## contrast estimate SE df lower.CL
## nas.non.chrom.IMI.cow1 - nas.non.chrom.IMI.cow0 0.0358 0.959 326 -1.850
## upper.CL
## 1.92
##
## postpartum_sample = 4:
## contrast estimate SE df lower.CL
## nas.non.chrom.IMI.cow1 - nas.non.chrom.IMI.cow0 0.4758 0.994 326 -1.480
## upper.CL
## 2.43
##
## postpartum_sample = 5:
## contrast estimate SE df lower.CL
## nas.non.chrom.IMI.cow1 - nas.non.chrom.IMI.cow0 1.4090 0.984 326 -0.526
## upper.CL
## 3.34
##
## postpartum_sample = 6:
## contrast estimate SE df lower.CL
## nas.non.chrom.IMI.cow1 - nas.non.chrom.IMI.cow0 2.0768 1.009 326 0.091
## upper.CL
## 4.06
##
## Results are averaged over the levels of: calving.sau, calving.sslo
## Degrees-of-freedom method: containment
## Confidence level used: 0.95adj_means_4 <- as.data.frame(emmeans_4$emmeans)
adj_means_4## postpartum_sample = 1:
## nas.non.chrom.IMI.cow emmean SE df lower.CL upper.CL
## 0 22.77726 3.155842 3 12.733965 32.82056
## 1 23.82080 3.251101 3 13.474348 34.16726
##
## postpartum_sample = 2:
## nas.non.chrom.IMI.cow emmean SE df lower.CL upper.CL
## 0 25.11265 3.156518 3 15.067204 35.15810
## 1 24.78396 3.245140 3 14.456477 35.11145
##
## postpartum_sample = 3:
## nas.non.chrom.IMI.cow emmean SE df lower.CL upper.CL
## 0 24.17942 3.158265 3 14.128417 34.23043
## 1 24.21522 3.248410 3 13.877328 34.55311
##
## postpartum_sample = 4:
## nas.non.chrom.IMI.cow emmean SE df lower.CL upper.CL
## 0 23.07374 3.165021 3 13.001236 33.14625
## 1 23.54955 3.254720 3 13.191578 33.90752
##
## postpartum_sample = 5:
## nas.non.chrom.IMI.cow emmean SE df lower.CL upper.CL
## 0 22.25951 3.161408 3 12.198503 32.32052
## 1 23.66853 3.253805 3 13.313474 34.02359
##
## postpartum_sample = 6:
## nas.non.chrom.IMI.cow emmean SE df lower.CL upper.CL
## 0 19.99689 3.162368 3 9.932821 30.06095
## 1 22.07368 3.261114 3 11.695364 32.45200
##
## Results are averaged over the levels of: calving.sau, calving.sslo
## Degrees-of-freedom method: containment
## Confidence level used: 0.95contrasts_4 <- as.data.frame(emmeans_4$contrasts) %>% select (postpartum_sample,p.value)
adj_means_4 <- adj_means_4 %>% left_join(contrasts_4, by="postpartum_sample")
adj_means_4 <- adj_means_4 %>% group_by(postpartum_sample) %>% mutate(max_milk = max(upper.CL)) %>% ungroup()
adj_means_4 <- adj_means_4 %>% mutate(p.value = ifelse(max_milk == upper.CL, p.value, 1))
adj_means_4$nas.non.chrom.IMI.cow <- as.factor(adj_means_4$nas.non.chrom.IMI.cow)
nas_nch_title <- expression(paste("NASM non-", italic("chromogenes")))
milk_4 <- ggplot(adj_means_4, aes(x=postpartum_sample, y=emmean, group=nas.non.chrom.IMI.cow, color=nas.non.chrom.IMI.cow)) +
geom_line() +
geom_point()+
geom_errorbar(aes(ymin=emmean-SE, ymax=emmean+SE), width=.2,
position=position_dodge(0.05)) +
ylim(0,32) +
labs(x="Monthly postpartum test",y="Milk production (Kg/day)",title=nas_nch_title,color="IMI") +
theme_classic() +
theme(plot.title=element_text(face="italic")) +
scale_color_viridis(discrete = TRUE) +
scale_color_manual(labels = c("No", "Yes"),
values = c("#009999", "#990000")) +
geom_text(aes(label = ifelse(p.value < 0.05, "*", "")),
vjust = -3, size = 5, color = "black")## Scale for colour is already present.
## Adding another scale for colour, which will replace the existing scale.model_4_adj_overall <- lme(fixed =milk_kg~ nas.non.chrom.IMI.cow+calving.sau+calving.sslo,random=~1|Site/CowId,correlation= corAR1(form = ~ps_sample),data=df.per.cow_dhi,na.action = na.omit)
Confint(model_4_adj_overall)## Estimate 2.5 % 97.5 %
## (Intercept) 23.8798870 17.7655051 29.9942690
## nas.non.chrom.IMI.cow1 0.8247931 -0.4734207 2.1230070
## calving.sau1 -0.0255785 -1.4427343 1.3915774
## calving.sslo1 -1.8850774 -3.1973404 -0.5728145# ---- Staph spp ----
model_5 <- lme(fixed =milk_kg~ staph.spp.IMI.cow*postpartum_sample+calving.sau+calving.sslo,random=~1|Site/CowId,correlation= corAR1(form = ~ps_sample),data=df.per.cow_dhi,na.action = na.omit)
emmeans(model_5,revpairwise~staph.spp.IMI.cow|postpartum_sample)## $emmeans
## postpartum_sample = 1:
## staph.spp.IMI.cow emmean SE df lower.CL upper.CL
## 0 22.9 3.15 3 12.85 32.9
## 1 23.5 3.26 3 13.17 33.9
##
## postpartum_sample = 2:
## staph.spp.IMI.cow emmean SE df lower.CL upper.CL
## 0 25.1 3.15 3 15.10 35.1
## 1 24.7 3.25 3 14.32 35.0
##
## postpartum_sample = 3:
## staph.spp.IMI.cow emmean SE df lower.CL upper.CL
## 0 24.2 3.15 3 14.17 34.2
## 1 24.1 3.25 3 13.76 34.5
##
## postpartum_sample = 4:
## staph.spp.IMI.cow emmean SE df lower.CL upper.CL
## 0 23.0 3.15 3 13.00 33.1
## 1 23.6 3.26 3 13.27 34.0
##
## postpartum_sample = 5:
## staph.spp.IMI.cow emmean SE df lower.CL upper.CL
## 0 22.3 3.15 3 12.25 32.3
## 1 23.7 3.26 3 13.35 34.1
##
## postpartum_sample = 6:
## staph.spp.IMI.cow emmean SE df lower.CL upper.CL
## 0 20.0 3.15 3 9.94 30.0
## 1 22.4 3.27 3 12.02 32.8
##
## Results are averaged over the levels of: calving.sau, calving.sslo
## Degrees-of-freedom method: containment
## Confidence level used: 0.95
##
## $contrasts
## postpartum_sample = 1:
## contrast estimate SE df t.ratio p.value
## staph.spp.IMI.cow1 - staph.spp.IMI.cow0 0.6702 1.009 326 0.664 0.5069
##
## postpartum_sample = 2:
## contrast estimate SE df t.ratio p.value
## staph.spp.IMI.cow1 - staph.spp.IMI.cow0 -0.4556 0.981 326 -0.464 0.6426
##
## postpartum_sample = 3:
## contrast estimate SE df t.ratio p.value
## staph.spp.IMI.cow1 - staph.spp.IMI.cow0 -0.0866 1.000 326 -0.087 0.9310
##
## postpartum_sample = 4:
## contrast estimate SE df t.ratio p.value
## staph.spp.IMI.cow1 - staph.spp.IMI.cow0 0.6061 1.038 326 0.584 0.5595
##
## postpartum_sample = 5:
## contrast estimate SE df t.ratio p.value
## staph.spp.IMI.cow1 - staph.spp.IMI.cow0 1.4437 1.028 326 1.405 0.1610
##
## postpartum_sample = 6:
## contrast estimate SE df t.ratio p.value
## staph.spp.IMI.cow1 - staph.spp.IMI.cow0 2.4508 1.053 326 2.328 0.0205
##
## Results are averaged over the levels of: calving.sau, calving.sslo
## Degrees-of-freedom method: containmentmodel_5 <- lme(fixed =milk_kg~ staph.spp.IMI.cow*postpartum_sample+calving.sau+calving.sslo,random=~1|Site/CowId,correlation= corAR1(form = ~ps_sample),data=df.per.cow_dhi,na.action = na.omit)
emmeans_5 <- emmeans(model_5,revpairwise~staph.spp.IMI.cow|postpartum_sample)
emmeans_5## $emmeans
## postpartum_sample = 1:
## staph.spp.IMI.cow emmean SE df lower.CL upper.CL
## 0 22.9 3.15 3 12.85 32.9
## 1 23.5 3.26 3 13.17 33.9
##
## postpartum_sample = 2:
## staph.spp.IMI.cow emmean SE df lower.CL upper.CL
## 0 25.1 3.15 3 15.10 35.1
## 1 24.7 3.25 3 14.32 35.0
##
## postpartum_sample = 3:
## staph.spp.IMI.cow emmean SE df lower.CL upper.CL
## 0 24.2 3.15 3 14.17 34.2
## 1 24.1 3.25 3 13.76 34.5
##
## postpartum_sample = 4:
## staph.spp.IMI.cow emmean SE df lower.CL upper.CL
## 0 23.0 3.15 3 13.00 33.1
## 1 23.6 3.26 3 13.27 34.0
##
## postpartum_sample = 5:
## staph.spp.IMI.cow emmean SE df lower.CL upper.CL
## 0 22.3 3.15 3 12.25 32.3
## 1 23.7 3.26 3 13.35 34.1
##
## postpartum_sample = 6:
## staph.spp.IMI.cow emmean SE df lower.CL upper.CL
## 0 20.0 3.15 3 9.94 30.0
## 1 22.4 3.27 3 12.02 32.8
##
## Results are averaged over the levels of: calving.sau, calving.sslo
## Degrees-of-freedom method: containment
## Confidence level used: 0.95
##
## $contrasts
## postpartum_sample = 1:
## contrast estimate SE df t.ratio p.value
## staph.spp.IMI.cow1 - staph.spp.IMI.cow0 0.6702 1.009 326 0.664 0.5069
##
## postpartum_sample = 2:
## contrast estimate SE df t.ratio p.value
## staph.spp.IMI.cow1 - staph.spp.IMI.cow0 -0.4556 0.981 326 -0.464 0.6426
##
## postpartum_sample = 3:
## contrast estimate SE df t.ratio p.value
## staph.spp.IMI.cow1 - staph.spp.IMI.cow0 -0.0866 1.000 326 -0.087 0.9310
##
## postpartum_sample = 4:
## contrast estimate SE df t.ratio p.value
## staph.spp.IMI.cow1 - staph.spp.IMI.cow0 0.6061 1.038 326 0.584 0.5595
##
## postpartum_sample = 5:
## contrast estimate SE df t.ratio p.value
## staph.spp.IMI.cow1 - staph.spp.IMI.cow0 1.4437 1.028 326 1.405 0.1610
##
## postpartum_sample = 6:
## contrast estimate SE df t.ratio p.value
## staph.spp.IMI.cow1 - staph.spp.IMI.cow0 2.4508 1.053 326 2.328 0.0205
##
## Results are averaged over the levels of: calving.sau, calving.sslo
## Degrees-of-freedom method: containmentconfint(emmeans_5)## $emmeans
## postpartum_sample = 1:
## staph.spp.IMI.cow emmean SE df lower.CL upper.CL
## 0 22.9 3.15 3 12.85 32.9
## 1 23.5 3.26 3 13.17 33.9
##
## postpartum_sample = 2:
## staph.spp.IMI.cow emmean SE df lower.CL upper.CL
## 0 25.1 3.15 3 15.10 35.1
## 1 24.7 3.25 3 14.32 35.0
##
## postpartum_sample = 3:
## staph.spp.IMI.cow emmean SE df lower.CL upper.CL
## 0 24.2 3.15 3 14.17 34.2
## 1 24.1 3.25 3 13.76 34.5
##
## postpartum_sample = 4:
## staph.spp.IMI.cow emmean SE df lower.CL upper.CL
## 0 23.0 3.15 3 13.00 33.1
## 1 23.6 3.26 3 13.27 34.0
##
## postpartum_sample = 5:
## staph.spp.IMI.cow emmean SE df lower.CL upper.CL
## 0 22.3 3.15 3 12.25 32.3
## 1 23.7 3.26 3 13.35 34.1
##
## postpartum_sample = 6:
## staph.spp.IMI.cow emmean SE df lower.CL upper.CL
## 0 20.0 3.15 3 9.94 30.0
## 1 22.4 3.27 3 12.02 32.8
##
## Results are averaged over the levels of: calving.sau, calving.sslo
## Degrees-of-freedom method: containment
## Confidence level used: 0.95
##
## $contrasts
## postpartum_sample = 1:
## contrast estimate SE df lower.CL upper.CL
## staph.spp.IMI.cow1 - staph.spp.IMI.cow0 0.6702 1.009 326 -1.314 2.65
##
## postpartum_sample = 2:
## contrast estimate SE df lower.CL upper.CL
## staph.spp.IMI.cow1 - staph.spp.IMI.cow0 -0.4556 0.981 326 -2.385 1.47
##
## postpartum_sample = 3:
## contrast estimate SE df lower.CL upper.CL
## staph.spp.IMI.cow1 - staph.spp.IMI.cow0 -0.0866 1.000 326 -2.053 1.88
##
## postpartum_sample = 4:
## contrast estimate SE df lower.CL upper.CL
## staph.spp.IMI.cow1 - staph.spp.IMI.cow0 0.6061 1.038 326 -1.435 2.65
##
## postpartum_sample = 5:
## contrast estimate SE df lower.CL upper.CL
## staph.spp.IMI.cow1 - staph.spp.IMI.cow0 1.4437 1.028 326 -0.578 3.47
##
## postpartum_sample = 6:
## contrast estimate SE df lower.CL upper.CL
## staph.spp.IMI.cow1 - staph.spp.IMI.cow0 2.4508 1.053 326 0.380 4.52
##
## Results are averaged over the levels of: calving.sau, calving.sslo
## Degrees-of-freedom method: containment
## Confidence level used: 0.95adj_means_5 <- as.data.frame(emmeans_5$emmeans)
adj_means_5## postpartum_sample = 1:
## staph.spp.IMI.cow emmean SE df lower.CL upper.CL
## 0 22.85510 3.145172 3 12.845754 32.86444
## 1 23.52531 3.255046 3 13.166303 33.88432
##
## postpartum_sample = 2:
## staph.spp.IMI.cow emmean SE df lower.CL upper.CL
## 0 25.11459 3.145831 3 15.103151 35.12603
## 1 24.65903 3.247230 3 14.324892 34.99316
##
## postpartum_sample = 3:
## staph.spp.IMI.cow emmean SE df lower.CL upper.CL
## 0 24.19024 3.147474 3 14.173570 34.20690
## 1 24.10364 3.251412 3 13.756198 34.45108
##
## postpartum_sample = 4:
## staph.spp.IMI.cow emmean SE df lower.CL upper.CL
## 0 23.04068 3.153831 3 13.003776 33.07757
## 1 23.64679 3.259443 3 13.273784 34.01979
##
## postpartum_sample = 5:
## staph.spp.IMI.cow emmean SE df lower.CL upper.CL
## 0 22.27919 3.150438 3 12.253085 32.30529
## 1 23.72287 3.258238 3 13.353705 34.09204
##
## postpartum_sample = 6:
## staph.spp.IMI.cow emmean SE df lower.CL upper.CL
## 0 19.96455 3.151442 3 9.935254 29.99385
## 1 22.41536 3.265752 3 12.022274 32.80844
##
## Results are averaged over the levels of: calving.sau, calving.sslo
## Degrees-of-freedom method: containment
## Confidence level used: 0.95contrasts_5 <- as.data.frame(emmeans_5$contrasts) %>% select (postpartum_sample,p.value)
adj_means_5 <- adj_means_5 %>% left_join(contrasts_5, by="postpartum_sample")
adj_means_5 <- adj_means_5 %>% group_by(postpartum_sample) %>% mutate(max_milk = max(upper.CL)) %>% ungroup()
adj_means_5 <- adj_means_5 %>% mutate(p.value = ifelse(max_milk == upper.CL, p.value, 1))
staph_sp_title <- expression(paste("Unspeciated ",italic("Staphylococcus ")))
milk_5 <- ggplot(adj_means_5, aes(x=postpartum_sample, y=emmean, group=staph.spp.IMI.cow, color=staph.spp.IMI.cow)) +
geom_line() +
geom_point()+
geom_errorbar(aes(ymin=emmean-SE, ymax=emmean+SE), width=.2,
position=position_dodge(0.05)) +
ylim(0,32) +
labs(x="Monthly postpartum test",y="Milk production (Kg/day)",title=staph_sp_title,color="IMI") +
theme_classic() +
theme(plot.title=element_text(face="italic")) +
scale_color_viridis(discrete = TRUE) +
scale_color_manual(labels = c("No", "Yes"),
values = c("#009999", "#990000")) +
geom_text(aes(label = ifelse(p.value < 0.05, "*", "")),
vjust = -3, size = 5, color = "black")## Scale for colour is already present.
## Adding another scale for colour, which will replace the existing scale.model_5_adj_overall <- lme(fixed =milk_kg~ staph.spp.IMI.cow+calving.sau+calving.sslo,random=~1|Site/CowId,correlation= corAR1(form = ~ps_sample),data=df.per.cow_dhi,na.action = na.omit)
Confint(model_5_adj_overall)## Estimate 2.5 % 97.5 %
## (Intercept) 23.91518652 17.8207897 30.0095833
## staph.spp.IMI.cow1 0.80355204 -0.5430581 2.1501622
## calving.sau1 -0.05030637 -1.4640500 1.3634372
## calving.sslo1 -1.90948946 -3.2209941 -0.5979848# ---- SSLO ----
model_6 <- lme(fixed =milk_kg~ strep.IMI.cow*postpartum_sample,random=~1|Site/CowId,correlation= corAR1(form = ~ps_sample),data=df.per.cow_dhi,na.action = na.omit)
emmeans(model_6,revpairwise~strep.IMI.cow|postpartum_sample)## $emmeans
## postpartum_sample = 1:
## strep.IMI.cow emmean SE df lower.CL upper.CL
## 0 24.0 3.16 3 13.9 34.0
## 1 22.8 3.19 3 12.6 32.9
##
## postpartum_sample = 2:
## strep.IMI.cow emmean SE df lower.CL upper.CL
## 0 26.4 3.16 3 16.3 36.5
## 1 24.2 3.19 3 14.0 34.3
##
## postpartum_sample = 3:
## strep.IMI.cow emmean SE df lower.CL upper.CL
## 0 25.2 3.17 3 15.1 35.2
## 1 23.9 3.20 3 13.7 34.1
##
## postpartum_sample = 4:
## strep.IMI.cow emmean SE df lower.CL upper.CL
## 0 24.2 3.17 3 14.1 34.3
## 1 22.9 3.21 3 12.7 33.1
##
## postpartum_sample = 5:
## strep.IMI.cow emmean SE df lower.CL upper.CL
## 0 23.3 3.17 3 13.2 33.3
## 1 22.9 3.20 3 12.7 33.1
##
## postpartum_sample = 6:
## strep.IMI.cow emmean SE df lower.CL upper.CL
## 0 21.2 3.17 3 11.1 31.3
## 1 20.6 3.21 3 10.4 30.8
##
## Degrees-of-freedom method: containment
## Confidence level used: 0.95
##
## $contrasts
## postpartum_sample = 1:
## contrast estimate SE df t.ratio p.value
## strep.IMI.cow1 - strep.IMI.cow0 -1.209 0.810 328 -1.493 0.1365
##
## postpartum_sample = 2:
## contrast estimate SE df t.ratio p.value
## strep.IMI.cow1 - strep.IMI.cow0 -2.249 0.815 328 -2.759 0.0061
##
## postpartum_sample = 3:
## contrast estimate SE df t.ratio p.value
## strep.IMI.cow1 - strep.IMI.cow0 -1.245 0.840 328 -1.483 0.1391
##
## postpartum_sample = 4:
## contrast estimate SE df t.ratio p.value
## strep.IMI.cow1 - strep.IMI.cow0 -1.234 0.884 328 -1.396 0.1637
##
## postpartum_sample = 5:
## contrast estimate SE df t.ratio p.value
## strep.IMI.cow1 - strep.IMI.cow0 -0.386 0.869 328 -0.445 0.6569
##
## postpartum_sample = 6:
## contrast estimate SE df t.ratio p.value
## strep.IMI.cow1 - strep.IMI.cow0 -0.649 0.886 328 -0.732 0.4646
##
## Degrees-of-freedom method: containmentmodel_6 <- lme(fixed =milk_kg~ strep.IMI.cow*postpartum_sample+calving.nas+calving.sau,random=~1|Site/CowId,correlation= corAR1(form = ~ps_sample),data=df.per.cow_dhi,na.action = na.omit)
emmeans_6 <- emmeans(model_6,revpairwise~strep.IMI.cow|postpartum_sample)
emmeans_6## $emmeans
## postpartum_sample = 1:
## strep.IMI.cow emmean SE df lower.CL upper.CL
## 0 23.9 3.16 3 13.8 33.9
## 1 22.6 3.19 3 12.5 32.8
##
## postpartum_sample = 2:
## strep.IMI.cow emmean SE df lower.CL upper.CL
## 0 26.3 3.16 3 16.3 36.4
## 1 24.0 3.19 3 13.8 34.2
##
## postpartum_sample = 3:
## strep.IMI.cow emmean SE df lower.CL upper.CL
## 0 25.1 3.16 3 15.0 35.1
## 1 23.8 3.20 3 13.6 34.0
##
## postpartum_sample = 4:
## strep.IMI.cow emmean SE df lower.CL upper.CL
## 0 24.1 3.17 3 14.0 34.2
## 1 22.8 3.20 3 12.6 33.0
##
## postpartum_sample = 5:
## strep.IMI.cow emmean SE df lower.CL upper.CL
## 0 23.2 3.16 3 13.1 33.2
## 1 22.7 3.20 3 12.5 32.9
##
## postpartum_sample = 6:
## strep.IMI.cow emmean SE df lower.CL upper.CL
## 0 21.1 3.17 3 11.1 31.2
## 1 20.4 3.21 3 10.2 30.6
##
## Results are averaged over the levels of: calving.nas, calving.sau
## Degrees-of-freedom method: containment
## Confidence level used: 0.95
##
## $contrasts
## postpartum_sample = 1:
## contrast estimate SE df t.ratio p.value
## strep.IMI.cow1 - strep.IMI.cow0 -1.263 0.817 326 -1.546 0.1232
##
## postpartum_sample = 2:
## contrast estimate SE df t.ratio p.value
## strep.IMI.cow1 - strep.IMI.cow0 -2.304 0.822 326 -2.801 0.0054
##
## postpartum_sample = 3:
## contrast estimate SE df t.ratio p.value
## strep.IMI.cow1 - strep.IMI.cow0 -1.299 0.847 326 -1.533 0.1262
##
## postpartum_sample = 4:
## contrast estimate SE df t.ratio p.value
## strep.IMI.cow1 - strep.IMI.cow0 -1.289 0.892 326 -1.445 0.1494
##
## postpartum_sample = 5:
## contrast estimate SE df t.ratio p.value
## strep.IMI.cow1 - strep.IMI.cow0 -0.440 0.876 326 -0.502 0.6157
##
## postpartum_sample = 6:
## contrast estimate SE df t.ratio p.value
## strep.IMI.cow1 - strep.IMI.cow0 -0.703 0.893 326 -0.787 0.4319
##
## Results are averaged over the levels of: calving.nas, calving.sau
## Degrees-of-freedom method: containmentconfint(emmeans_6)## $emmeans
## postpartum_sample = 1:
## strep.IMI.cow emmean SE df lower.CL upper.CL
## 0 23.9 3.16 3 13.8 33.9
## 1 22.6 3.19 3 12.5 32.8
##
## postpartum_sample = 2:
## strep.IMI.cow emmean SE df lower.CL upper.CL
## 0 26.3 3.16 3 16.3 36.4
## 1 24.0 3.19 3 13.8 34.2
##
## postpartum_sample = 3:
## strep.IMI.cow emmean SE df lower.CL upper.CL
## 0 25.1 3.16 3 15.0 35.1
## 1 23.8 3.20 3 13.6 34.0
##
## postpartum_sample = 4:
## strep.IMI.cow emmean SE df lower.CL upper.CL
## 0 24.1 3.17 3 14.0 34.2
## 1 22.8 3.20 3 12.6 33.0
##
## postpartum_sample = 5:
## strep.IMI.cow emmean SE df lower.CL upper.CL
## 0 23.2 3.16 3 13.1 33.2
## 1 22.7 3.20 3 12.5 32.9
##
## postpartum_sample = 6:
## strep.IMI.cow emmean SE df lower.CL upper.CL
## 0 21.1 3.17 3 11.1 31.2
## 1 20.4 3.21 3 10.2 30.6
##
## Results are averaged over the levels of: calving.nas, calving.sau
## Degrees-of-freedom method: containment
## Confidence level used: 0.95
##
## $contrasts
## postpartum_sample = 1:
## contrast estimate SE df lower.CL upper.CL
## strep.IMI.cow1 - strep.IMI.cow0 -1.263 0.817 326 -2.87 0.344
##
## postpartum_sample = 2:
## contrast estimate SE df lower.CL upper.CL
## strep.IMI.cow1 - strep.IMI.cow0 -2.304 0.822 326 -3.92 -0.686
##
## postpartum_sample = 3:
## contrast estimate SE df lower.CL upper.CL
## strep.IMI.cow1 - strep.IMI.cow0 -1.299 0.847 326 -2.97 0.368
##
## postpartum_sample = 4:
## contrast estimate SE df lower.CL upper.CL
## strep.IMI.cow1 - strep.IMI.cow0 -1.289 0.892 326 -3.04 0.465
##
## postpartum_sample = 5:
## contrast estimate SE df lower.CL upper.CL
## strep.IMI.cow1 - strep.IMI.cow0 -0.440 0.876 326 -2.16 1.284
##
## postpartum_sample = 6:
## contrast estimate SE df lower.CL upper.CL
## strep.IMI.cow1 - strep.IMI.cow0 -0.703 0.893 326 -2.46 1.055
##
## Results are averaged over the levels of: calving.nas, calving.sau
## Degrees-of-freedom method: containment
## Confidence level used: 0.95adj_means_6 <- as.data.frame(emmeans_6$emmeans)
adj_means_6## postpartum_sample = 1:
## strep.IMI.cow emmean SE df lower.CL upper.CL
## 0 23.88439 3.159264 3 13.83020 33.93858
## 1 22.62176 3.191330 3 12.46552 32.77799
##
## postpartum_sample = 2:
## strep.IMI.cow emmean SE df lower.CL upper.CL
## 0 26.31171 3.158844 3 16.25886 36.36456
## 1 24.00801 3.193476 3 13.84494 34.17108
##
## postpartum_sample = 3:
## strep.IMI.cow emmean SE df lower.CL upper.CL
## 0 25.08200 3.160757 3 15.02307 35.14094
## 1 23.78295 3.197742 3 13.60631 33.95959
##
## postpartum_sample = 4:
## strep.IMI.cow emmean SE df lower.CL upper.CL
## 0 24.06555 3.168931 3 13.98060 34.15051
## 1 22.77702 3.203989 3 12.58050 32.97355
##
## postpartum_sample = 5:
## strep.IMI.cow emmean SE df lower.CL upper.CL
## 0 23.16981 3.164363 3 13.09939 33.24022
## 1 22.72948 3.202854 3 12.53656 32.92239
##
## postpartum_sample = 6:
## strep.IMI.cow emmean SE df lower.CL upper.CL
## 0 21.14676 3.165668 3 11.07219 31.22133
## 1 20.44382 3.206366 3 10.23973 30.64791
##
## Results are averaged over the levels of: calving.nas, calving.sau
## Degrees-of-freedom method: containment
## Confidence level used: 0.95contrasts_6 <- as.data.frame(emmeans_6$contrasts) %>% select (postpartum_sample,p.value)
adj_means_6 <- adj_means_6 %>% left_join(contrasts_6, by="postpartum_sample")
adj_means_6 <- adj_means_6 %>% group_by(postpartum_sample) %>% mutate(max_milk = max(upper.CL)) %>% ungroup()
adj_means_6 <- adj_means_6 %>% mutate(p.value = ifelse(max_milk == upper.CL, p.value, 1))
#staph_sp_title <- expression(paste(italic("Staphylococcus "), "sp."))
#sslo_title <- expression(paste(italic("Streptococcus "), "spp. and ",italic("Streptococcus"),"-like organisms"))
milk_6 <- ggplot(adj_means_6, aes(x=postpartum_sample, y=emmean, group=strep.IMI.cow, color=strep.IMI.cow)) +
geom_line() +
geom_point()+
geom_errorbar(aes(ymin=emmean-SE, ymax=emmean+SE), width=.2,
position=position_dodge(0.05)) +
ylim(0,32) +
labs(x="Monthly postpartum test",y="Milk production (Kg/day)",title="SSLO",color="IMI") +
theme_classic() +
scale_color_viridis(discrete = TRUE) +
scale_color_manual(labels = c("No", "Yes"),
values = c("#009999", "#990000")) +
geom_text(aes(label = ifelse(p.value < 0.05, "*", "")),
vjust = -3, size = 5, color = "black")## Scale for colour is already present.
## Adding another scale for colour, which will replace the existing scale.model_6_adj_overall <- lme(fixed =milk_kg~ strep.IMI.cow+calving.sau+calving.nas,random=~1|Site/CowId,correlation= corAR1(form = ~ps_sample),data=df.per.cow_dhi,na.action = na.omit)
Confint(model_6_adj_overall)## Estimate 2.5 % 97.5 %
## (Intercept) 24.1913418 18.061248 30.32143511
## strep.IMI.cow1 -1.1965434 -2.347216 -0.04587109
## calving.sau1 -0.1142546 -1.542970 1.31446042
## calving.nas1 -0.3249312 -1.538226 0.88836401# ---- Streptococcus spp ----
model_7 <- lme(fixed =milk_kg~ strep.genus.IMI.cow*postpartum_sample,random=~1|Site/CowId,correlation= corAR1(form = ~ps_sample),data=df.per.cow_dhi,na.action = na.omit)
emmeans(model_7,revpairwise~strep.genus.IMI.cow|postpartum_sample)## $emmeans
## postpartum_sample = 1:
## strep.genus.IMI.cow emmean SE df lower.CL upper.CL
## 0 23.8 3.15 3 13.7 33.8
## 1 22.8 3.24 3 12.4 33.1
##
## postpartum_sample = 2:
## strep.genus.IMI.cow emmean SE df lower.CL upper.CL
## 0 26.1 3.15 3 16.1 36.2
## 1 23.6 3.24 3 13.3 33.9
##
## postpartum_sample = 3:
## strep.genus.IMI.cow emmean SE df lower.CL upper.CL
## 0 25.0 3.16 3 15.0 35.1
## 1 23.7 3.25 3 13.3 34.0
##
## postpartum_sample = 4:
## strep.genus.IMI.cow emmean SE df lower.CL upper.CL
## 0 24.0 3.16 3 13.9 34.1
## 1 22.6 3.26 3 12.3 33.0
##
## postpartum_sample = 5:
## strep.genus.IMI.cow emmean SE df lower.CL upper.CL
## 0 23.0 3.16 3 13.0 33.1
## 1 23.5 3.26 3 13.1 33.8
##
## postpartum_sample = 6:
## strep.genus.IMI.cow emmean SE df lower.CL upper.CL
## 0 21.1 3.16 3 11.0 31.1
## 1 20.7 3.26 3 10.3 31.1
##
## Degrees-of-freedom method: containment
## Confidence level used: 0.95
##
## $contrasts
## postpartum_sample = 1:
## contrast estimate SE df t.ratio p.value
## strep.genus.IMI.cow1 - strep.genus.IMI.cow0 -0.997 0.979 328 -1.019 0.3092
##
## postpartum_sample = 2:
## contrast estimate SE df t.ratio p.value
## strep.genus.IMI.cow1 - strep.genus.IMI.cow0 -2.514 0.980 328 -2.564 0.0108
##
## postpartum_sample = 3:
## contrast estimate SE df t.ratio p.value
## strep.genus.IMI.cow1 - strep.genus.IMI.cow0 -1.342 1.006 328 -1.334 0.1832
##
## postpartum_sample = 4:
## contrast estimate SE df t.ratio p.value
## strep.genus.IMI.cow1 - strep.genus.IMI.cow0 -1.366 1.046 328 -1.306 0.1925
##
## postpartum_sample = 5:
## contrast estimate SE df t.ratio p.value
## strep.genus.IMI.cow1 - strep.genus.IMI.cow0 0.422 1.035 328 0.408 0.6838
##
## postpartum_sample = 6:
## contrast estimate SE df t.ratio p.value
## strep.genus.IMI.cow1 - strep.genus.IMI.cow0 -0.386 1.044 328 -0.370 0.7119
##
## Degrees-of-freedom method: containmentmodel_7 <- lme(fixed =milk_kg~ strep.genus.IMI.cow*postpartum_sample+calving.nas+calving.sau,random=~1|Site/CowId,correlation= corAR1(form = ~ps_sample),data=df.per.cow_dhi,na.action = na.omit)
emmeans_7 <- emmeans(model_7,revpairwise~strep.genus.IMI.cow|postpartum_sample)
emmeans_7## $emmeans
## postpartum_sample = 1:
## strep.genus.IMI.cow emmean SE df lower.CL upper.CL
## 0 23.7 3.16 3 13.6 33.7
## 1 22.7 3.25 3 12.3 33.0
##
## postpartum_sample = 2:
## strep.genus.IMI.cow emmean SE df lower.CL upper.CL
## 0 26.1 3.16 3 16.0 36.1
## 1 23.5 3.25 3 13.2 33.8
##
## postpartum_sample = 3:
## strep.genus.IMI.cow emmean SE df lower.CL upper.CL
## 0 25.0 3.16 3 14.9 35.0
## 1 23.6 3.25 3 13.2 33.9
##
## postpartum_sample = 4:
## strep.genus.IMI.cow emmean SE df lower.CL upper.CL
## 0 24.0 3.17 3 13.9 34.0
## 1 22.5 3.26 3 12.2 32.9
##
## postpartum_sample = 5:
## strep.genus.IMI.cow emmean SE df lower.CL upper.CL
## 0 23.0 3.16 3 12.9 33.0
## 1 23.4 3.26 3 13.0 33.7
##
## postpartum_sample = 6:
## strep.genus.IMI.cow emmean SE df lower.CL upper.CL
## 0 21.0 3.16 3 11.0 31.1
## 1 20.6 3.26 3 10.2 31.0
##
## Results are averaged over the levels of: calving.nas, calving.sau
## Degrees-of-freedom method: containment
## Confidence level used: 0.95
##
## $contrasts
## postpartum_sample = 1:
## contrast estimate SE df t.ratio p.value
## strep.genus.IMI.cow1 - strep.genus.IMI.cow0 -1.037 0.988 326 -1.049 0.2948
##
## postpartum_sample = 2:
## contrast estimate SE df t.ratio p.value
## strep.genus.IMI.cow1 - strep.genus.IMI.cow0 -2.554 0.989 326 -2.582 0.0103
##
## postpartum_sample = 3:
## contrast estimate SE df t.ratio p.value
## strep.genus.IMI.cow1 - strep.genus.IMI.cow0 -1.382 1.015 326 -1.361 0.1743
##
## postpartum_sample = 4:
## contrast estimate SE df t.ratio p.value
## strep.genus.IMI.cow1 - strep.genus.IMI.cow0 -1.406 1.055 326 -1.333 0.1835
##
## postpartum_sample = 5:
## contrast estimate SE df t.ratio p.value
## strep.genus.IMI.cow1 - strep.genus.IMI.cow0 0.382 1.044 326 0.366 0.7149
##
## postpartum_sample = 6:
## contrast estimate SE df t.ratio p.value
## strep.genus.IMI.cow1 - strep.genus.IMI.cow0 -0.426 1.053 326 -0.405 0.6860
##
## Results are averaged over the levels of: calving.nas, calving.sau
## Degrees-of-freedom method: containmentconfint(emmeans_7)## $emmeans
## postpartum_sample = 1:
## strep.genus.IMI.cow emmean SE df lower.CL upper.CL
## 0 23.7 3.16 3 13.6 33.7
## 1 22.7 3.25 3 12.3 33.0
##
## postpartum_sample = 2:
## strep.genus.IMI.cow emmean SE df lower.CL upper.CL
## 0 26.1 3.16 3 16.0 36.1
## 1 23.5 3.25 3 13.2 33.8
##
## postpartum_sample = 3:
## strep.genus.IMI.cow emmean SE df lower.CL upper.CL
## 0 25.0 3.16 3 14.9 35.0
## 1 23.6 3.25 3 13.2 33.9
##
## postpartum_sample = 4:
## strep.genus.IMI.cow emmean SE df lower.CL upper.CL
## 0 24.0 3.17 3 13.9 34.0
## 1 22.5 3.26 3 12.2 32.9
##
## postpartum_sample = 5:
## strep.genus.IMI.cow emmean SE df lower.CL upper.CL
## 0 23.0 3.16 3 12.9 33.0
## 1 23.4 3.26 3 13.0 33.7
##
## postpartum_sample = 6:
## strep.genus.IMI.cow emmean SE df lower.CL upper.CL
## 0 21.0 3.16 3 11.0 31.1
## 1 20.6 3.26 3 10.2 31.0
##
## Results are averaged over the levels of: calving.nas, calving.sau
## Degrees-of-freedom method: containment
## Confidence level used: 0.95
##
## $contrasts
## postpartum_sample = 1:
## contrast estimate SE df lower.CL
## strep.genus.IMI.cow1 - strep.genus.IMI.cow0 -1.037 0.988 326 -2.98
## upper.CL
## 0.907
##
## postpartum_sample = 2:
## contrast estimate SE df lower.CL
## strep.genus.IMI.cow1 - strep.genus.IMI.cow0 -2.554 0.989 326 -4.50
## upper.CL
## -0.608
##
## postpartum_sample = 3:
## contrast estimate SE df lower.CL
## strep.genus.IMI.cow1 - strep.genus.IMI.cow0 -1.382 1.015 326 -3.38
## upper.CL
## 0.615
##
## postpartum_sample = 4:
## contrast estimate SE df lower.CL
## strep.genus.IMI.cow1 - strep.genus.IMI.cow0 -1.406 1.055 326 -3.48
## upper.CL
## 0.669
##
## postpartum_sample = 5:
## contrast estimate SE df lower.CL
## strep.genus.IMI.cow1 - strep.genus.IMI.cow0 0.382 1.044 326 -1.67
## upper.CL
## 2.437
##
## postpartum_sample = 6:
## contrast estimate SE df lower.CL
## strep.genus.IMI.cow1 - strep.genus.IMI.cow0 -0.426 1.053 326 -2.50
## upper.CL
## 1.646
##
## Results are averaged over the levels of: calving.nas, calving.sau
## Degrees-of-freedom method: containment
## Confidence level used: 0.95adj_means_7 <- as.data.frame(emmeans_7$emmeans)
adj_means_7## postpartum_sample = 1:
## strep.genus.IMI.cow emmean SE df lower.CL upper.CL
## 0 23.69725 3.157400 3 13.64899 33.74550
## 1 22.66040 3.246932 3 12.32721 32.99359
##
## postpartum_sample = 2:
## strep.genus.IMI.cow emmean SE df lower.CL upper.CL
## 0 26.05944 3.157444 3 16.01104 36.10783
## 1 23.50559 3.247803 3 13.16963 33.84155
##
## postpartum_sample = 3:
## strep.genus.IMI.cow emmean SE df lower.CL upper.CL
## 0 24.95237 3.159185 3 14.89843 35.00630
## 1 23.57075 3.253524 3 13.21659 33.92492
##
## postpartum_sample = 4:
## strep.genus.IMI.cow emmean SE df lower.CL upper.CL
## 0 23.95306 3.165449 3 13.87919 34.02693
## 1 22.54710 3.262135 3 12.16553 32.92867
##
## postpartum_sample = 5:
## strep.genus.IMI.cow emmean SE df lower.CL upper.CL
## 0 22.97878 3.162177 3 12.91532 33.04224
## 1 23.36067 3.260727 3 12.98358 33.73776
##
## postpartum_sample = 6:
## strep.genus.IMI.cow emmean SE df lower.CL upper.CL
## 0 21.03048 3.163562 3 10.96261 31.09834
## 1 20.60435 3.262575 3 10.22138 30.98732
##
## Results are averaged over the levels of: calving.nas, calving.sau
## Degrees-of-freedom method: containment
## Confidence level used: 0.95contrasts_7 <- as.data.frame(emmeans_7$contrasts) %>% select (postpartum_sample,p.value)
adj_means_7 <- adj_means_7 %>% left_join(contrasts_7, by="postpartum_sample")
adj_means_7 <- adj_means_7 %>% group_by(postpartum_sample) %>% mutate(max_milk = max(upper.CL)) %>% ungroup()
adj_means_7 <- adj_means_7 %>% mutate(p.value = ifelse(max_milk == upper.CL, p.value, 1))
strep_genus_title <- expression(paste(italic("Streptococcus "), "spp."))
milk_7 <- ggplot(adj_means_7, aes(x=postpartum_sample, y=emmean, group=strep.genus.IMI.cow, color=strep.genus.IMI.cow)) +
geom_line() +
geom_point()+
geom_errorbar(aes(ymin=emmean-SE, ymax=emmean+SE), width=.2,
position=position_dodge(0.05)) +
ylim(0,32) +
labs(x="Monthly postpartum test",y="Milk production (Kg/day)",title=strep_genus_title,color="IMI") +
theme_classic() +
theme(plot.title=element_text(face="italic")) +
scale_color_viridis(discrete = TRUE) +
scale_color_manual(labels = c("No", "Yes"),
values = c("#009999", "#990000")) +
geom_text(aes(label = ifelse(p.value < 0.05, "*", "")),
vjust = -3, size = 5, color = "black")## Scale for colour is already present.
## Adding another scale for colour, which will replace the existing scale.model_7_adj_overall <- lme(fixed =milk_kg~ strep.genus.IMI.cow+calving.sau+calving.nas,random=~1|Site/CowId,correlation= corAR1(form = ~ps_sample),data=df.per.cow_dhi,na.action = na.omit)
Confint(model_7_adj_overall)## Estimate 2.5 % 97.5 %
## (Intercept) 23.95765291 17.834451 30.0808549
## strep.genus.IMI.cow1 -1.10822834 -2.505762 0.2893049
## calving.sau1 -0.03882997 -1.475496 1.3978356
## calving.nas1 -0.23813552 -1.446676 0.9704050# ----Streptococcus dysgalactiae ----
model_8 <- lme(fixed =milk_kg ~ strep.dys.IMI.cow*postpartum_sample,random=~1|Site/CowId,correlation= corAR1(form = ~ps_sample),data=df.per.cow_dhi,na.action = na.omit)
emmeans(model_8,revpairwise~strep.dys.IMI.cow|postpartum_sample)## $emmeans
## postpartum_sample = 1:
## strep.dys.IMI.cow emmean SE df lower.CL upper.CL
## 0 23.9 3.18 3 13.8 34.0
## 1 21.2 3.34 3 10.6 31.9
##
## postpartum_sample = 2:
## strep.dys.IMI.cow emmean SE df lower.CL upper.CL
## 0 26.0 3.18 3 15.9 36.2
## 1 22.5 3.33 3 11.9 33.1
##
## postpartum_sample = 3:
## strep.dys.IMI.cow emmean SE df lower.CL upper.CL
## 0 24.9 3.18 3 14.8 35.1
## 1 23.3 3.34 3 12.7 34.0
##
## postpartum_sample = 4:
## strep.dys.IMI.cow emmean SE df lower.CL upper.CL
## 0 23.7 3.19 3 13.6 33.9
## 1 23.3 3.36 3 12.6 34.0
##
## postpartum_sample = 5:
## strep.dys.IMI.cow emmean SE df lower.CL upper.CL
## 0 23.1 3.18 3 12.9 33.2
## 1 23.3 3.35 3 12.6 33.9
##
## postpartum_sample = 6:
## strep.dys.IMI.cow emmean SE df lower.CL upper.CL
## 0 21.0 3.19 3 10.8 31.1
## 1 21.1 3.35 3 10.4 31.8
##
## Degrees-of-freedom method: containment
## Confidence level used: 0.95
##
## $contrasts
## postpartum_sample = 1:
## contrast estimate SE df t.ratio p.value
## strep.dys.IMI.cow1 - strep.dys.IMI.cow0 -2.633 1.16 328 -2.264 0.0243
##
## postpartum_sample = 2:
## contrast estimate SE df t.ratio p.value
## strep.dys.IMI.cow1 - strep.dys.IMI.cow0 -3.528 1.15 328 -3.054 0.0024
##
## postpartum_sample = 3:
## contrast estimate SE df t.ratio p.value
## strep.dys.IMI.cow1 - strep.dys.IMI.cow0 -1.584 1.18 328 -1.344 0.1799
##
## postpartum_sample = 4:
## contrast estimate SE df t.ratio p.value
## strep.dys.IMI.cow1 - strep.dys.IMI.cow0 -0.427 1.23 328 -0.347 0.7286
##
## postpartum_sample = 5:
## contrast estimate SE df t.ratio p.value
## strep.dys.IMI.cow1 - strep.dys.IMI.cow0 0.184 1.22 328 0.151 0.8800
##
## postpartum_sample = 6:
## contrast estimate SE df t.ratio p.value
## strep.dys.IMI.cow1 - strep.dys.IMI.cow0 0.153 1.22 328 0.126 0.8999
##
## Degrees-of-freedom method: containmentmodel_8 <- lme(fixed =milk_kg ~ strep.dys.IMI.cow*postpartum_sample+calving.nas+calving.sau,random=~1|Site/CowId,correlation= corAR1(form = ~ps_sample),data=df.per.cow_dhi,na.action = na.omit)
emmeans_8 <- emmeans(model_8,revpairwise~strep.dys.IMI.cow|postpartum_sample)
emmeans_8## $emmeans
## postpartum_sample = 1:
## strep.dys.IMI.cow emmean SE df lower.CL upper.CL
## 0 23.8 3.18 3 13.7 33.9
## 1 21.1 3.34 3 10.5 31.7
##
## postpartum_sample = 2:
## strep.dys.IMI.cow emmean SE df lower.CL upper.CL
## 0 26.0 3.18 3 15.8 36.1
## 1 22.4 3.34 3 11.8 33.0
##
## postpartum_sample = 3:
## strep.dys.IMI.cow emmean SE df lower.CL upper.CL
## 0 24.8 3.18 3 14.7 34.9
## 1 23.2 3.34 3 12.6 33.8
##
## postpartum_sample = 4:
## strep.dys.IMI.cow emmean SE df lower.CL upper.CL
## 0 23.6 3.19 3 13.5 33.8
## 1 23.2 3.36 3 12.5 33.9
##
## postpartum_sample = 5:
## strep.dys.IMI.cow emmean SE df lower.CL upper.CL
## 0 23.0 3.18 3 12.9 33.1
## 1 23.1 3.36 3 12.4 33.8
##
## postpartum_sample = 6:
## strep.dys.IMI.cow emmean SE df lower.CL upper.CL
## 0 20.9 3.18 3 10.7 31.0
## 1 21.0 3.36 3 10.3 31.7
##
## Results are averaged over the levels of: calving.nas, calving.sau
## Degrees-of-freedom method: containment
## Confidence level used: 0.95
##
## $contrasts
## postpartum_sample = 1:
## contrast estimate SE df t.ratio p.value
## strep.dys.IMI.cow1 - strep.dys.IMI.cow0 -2.674 1.17 326 -2.287 0.0228
##
## postpartum_sample = 2:
## contrast estimate SE df t.ratio p.value
## strep.dys.IMI.cow1 - strep.dys.IMI.cow0 -3.571 1.16 326 -3.074 0.0023
##
## postpartum_sample = 3:
## contrast estimate SE df t.ratio p.value
## strep.dys.IMI.cow1 - strep.dys.IMI.cow0 -1.625 1.19 326 -1.371 0.1713
##
## postpartum_sample = 4:
## contrast estimate SE df t.ratio p.value
## strep.dys.IMI.cow1 - strep.dys.IMI.cow0 -0.468 1.24 326 -0.378 0.7054
##
## postpartum_sample = 5:
## contrast estimate SE df t.ratio p.value
## strep.dys.IMI.cow1 - strep.dys.IMI.cow0 0.143 1.22 326 0.117 0.9071
##
## postpartum_sample = 6:
## contrast estimate SE df t.ratio p.value
## strep.dys.IMI.cow1 - strep.dys.IMI.cow0 0.114 1.23 326 0.093 0.9257
##
## Results are averaged over the levels of: calving.nas, calving.sau
## Degrees-of-freedom method: containmentconfint(emmeans_8)## $emmeans
## postpartum_sample = 1:
## strep.dys.IMI.cow emmean SE df lower.CL upper.CL
## 0 23.8 3.18 3 13.7 33.9
## 1 21.1 3.34 3 10.5 31.7
##
## postpartum_sample = 2:
## strep.dys.IMI.cow emmean SE df lower.CL upper.CL
## 0 26.0 3.18 3 15.8 36.1
## 1 22.4 3.34 3 11.8 33.0
##
## postpartum_sample = 3:
## strep.dys.IMI.cow emmean SE df lower.CL upper.CL
## 0 24.8 3.18 3 14.7 34.9
## 1 23.2 3.34 3 12.6 33.8
##
## postpartum_sample = 4:
## strep.dys.IMI.cow emmean SE df lower.CL upper.CL
## 0 23.6 3.19 3 13.5 33.8
## 1 23.2 3.36 3 12.5 33.9
##
## postpartum_sample = 5:
## strep.dys.IMI.cow emmean SE df lower.CL upper.CL
## 0 23.0 3.18 3 12.9 33.1
## 1 23.1 3.36 3 12.4 33.8
##
## postpartum_sample = 6:
## strep.dys.IMI.cow emmean SE df lower.CL upper.CL
## 0 20.9 3.18 3 10.7 31.0
## 1 21.0 3.36 3 10.3 31.7
##
## Results are averaged over the levels of: calving.nas, calving.sau
## Degrees-of-freedom method: containment
## Confidence level used: 0.95
##
## $contrasts
## postpartum_sample = 1:
## contrast estimate SE df lower.CL upper.CL
## strep.dys.IMI.cow1 - strep.dys.IMI.cow0 -2.674 1.17 326 -4.97 -0.374
##
## postpartum_sample = 2:
## contrast estimate SE df lower.CL upper.CL
## strep.dys.IMI.cow1 - strep.dys.IMI.cow0 -3.571 1.16 326 -5.86 -1.286
##
## postpartum_sample = 3:
## contrast estimate SE df lower.CL upper.CL
## strep.dys.IMI.cow1 - strep.dys.IMI.cow0 -1.625 1.19 326 -3.96 0.707
##
## postpartum_sample = 4:
## contrast estimate SE df lower.CL upper.CL
## strep.dys.IMI.cow1 - strep.dys.IMI.cow0 -0.468 1.24 326 -2.90 1.963
##
## postpartum_sample = 5:
## contrast estimate SE df lower.CL upper.CL
## strep.dys.IMI.cow1 - strep.dys.IMI.cow0 0.143 1.22 326 -2.26 2.547
##
## postpartum_sample = 6:
## contrast estimate SE df lower.CL upper.CL
## strep.dys.IMI.cow1 - strep.dys.IMI.cow0 0.114 1.23 326 -2.30 2.525
##
## Results are averaged over the levels of: calving.nas, calving.sau
## Degrees-of-freedom method: containment
## Confidence level used: 0.95adj_means_8 <- as.data.frame(emmeans_8$emmeans)
adj_means_8## postpartum_sample = 1:
## strep.dys.IMI.cow emmean SE df lower.CL upper.CL
## 0 23.78478 3.179046 3 13.66764 33.90193
## 1 21.11050 3.339727 3 10.48200 31.73900
##
## postpartum_sample = 2:
## strep.dys.IMI.cow emmean SE df lower.CL upper.CL
## 0 25.95271 3.179171 3 15.83517 36.07025
## 1 22.38156 3.337909 3 11.75884 33.00428
##
## postpartum_sample = 3:
## strep.dys.IMI.cow emmean SE df lower.CL upper.CL
## 0 24.82742 3.180732 3 14.70491 34.94993
## 1 23.20194 3.344357 3 12.55871 33.84518
##
## postpartum_sample = 4:
## strep.dys.IMI.cow emmean SE df lower.CL upper.CL
## 0 23.64398 3.186213 3 13.50403 33.78393
## 1 23.17645 3.358895 3 12.48695 33.86595
##
## postpartum_sample = 5:
## strep.dys.IMI.cow emmean SE df lower.CL upper.CL
## 0 22.98336 3.183433 3 12.85225 33.11446
## 1 23.12611 3.355407 3 12.44771 33.80451
##
## postpartum_sample = 6:
## strep.dys.IMI.cow emmean SE df lower.CL upper.CL
## 0 20.86260 3.184700 3 10.72746 30.99773
## 1 20.97700 3.355789 3 10.29738 31.65662
##
## Results are averaged over the levels of: calving.nas, calving.sau
## Degrees-of-freedom method: containment
## Confidence level used: 0.95contrasts_8 <- as.data.frame(emmeans_8$contrasts) %>% select (postpartum_sample,p.value)
adj_means_8 <- adj_means_8 %>% left_join(contrasts_8, by="postpartum_sample")
adj_means_8 <- adj_means_8 %>% group_by(postpartum_sample) %>% mutate(max_milk = max(upper.CL)) %>% ungroup()
adj_means_8 <- adj_means_8 %>% mutate(p.value = ifelse(max_milk == upper.CL, p.value, 1))
staph_sp_title <- expression(paste(italic("Staphylococcus "), "sp."))
milk_8 <- ggplot(adj_means_8, aes(x=postpartum_sample, y=emmean, group=strep.dys.IMI.cow, color=strep.dys.IMI.cow)) +
geom_line() +
geom_point()+
geom_errorbar(aes(ymin=emmean-SE, ymax=emmean+SE), width=.2,
position=position_dodge(0.05)) +
ylim(0,32) +
labs(x="Monthly postpartum test",y="Milk production (Kg/day)",title="Streptococcus dysgalactiae",color="IMI") +
theme_classic() +
theme(plot.title=element_text(face="italic")) +
scale_color_viridis(discrete = TRUE) +
scale_color_manual(labels = c("No", "Yes"),
values = c("#009999", "#990000")) +
geom_text(aes(label = ifelse(p.value < 0.05, "*", "")),
vjust = -3, size = 5, color = "black")## Scale for colour is already present.
## Adding another scale for colour, which will replace the existing scale.model_8_adj_overall <- lme(fixed =milk_kg~ strep.dys.IMI.cow+calving.sau+calving.nas,random=~1|Site/CowId,correlation= corAR1(form = ~ps_sample),data=df.per.cow_dhi,na.action = na.omit)
Confint(model_8_adj_overall)## Estimate 2.5 % 97.5 %
## (Intercept) 23.9332044 17.755659 30.1107500
## strep.dys.IMI.cow1 -1.4750149 -3.088485 0.1384552
## calving.sau1 -0.1215295 -1.551452 1.3083932
## calving.nas1 -0.2490436 -1.455491 0.9574034# ---- Unspeciated Strep spp. ----
model_9 <- lme(fixed =milk_kg ~ strep.sp.IMI.cow*postpartum_sample,random=~1|Site/CowId,correlation= corAR1(form = ~ps_sample),data=df.per.cow_dhi,na.action = na.omit)
emmeans(model_9,revpairwise~strep.sp.IMI.cow|postpartum_sample)## $emmeans
## postpartum_sample = 1:
## strep.sp.IMI.cow emmean SE df lower.CL upper.CL
## 0 23.6 3.18 3 13.5 33.7
## 1 23.0 3.40 3 12.1 33.8
##
## postpartum_sample = 2:
## strep.sp.IMI.cow emmean SE df lower.CL upper.CL
## 0 25.8 3.18 3 15.7 36.0
## 1 23.2 3.40 3 12.4 34.0
##
## postpartum_sample = 3:
## strep.sp.IMI.cow emmean SE df lower.CL upper.CL
## 0 24.9 3.19 3 14.7 35.0
## 1 23.5 3.41 3 12.7 34.4
##
## postpartum_sample = 4:
## strep.sp.IMI.cow emmean SE df lower.CL upper.CL
## 0 24.0 3.19 3 13.9 34.2
## 1 21.2 3.42 3 10.3 32.1
##
## postpartum_sample = 5:
## strep.sp.IMI.cow emmean SE df lower.CL upper.CL
## 0 23.1 3.19 3 13.0 33.2
## 1 23.4 3.43 3 12.5 34.3
##
## postpartum_sample = 6:
## strep.sp.IMI.cow emmean SE df lower.CL upper.CL
## 0 21.0 3.19 3 10.9 31.2
## 1 20.7 3.43 3 9.8 31.6
##
## Degrees-of-freedom method: containment
## Confidence level used: 0.95
##
## $contrasts
## postpartum_sample = 1:
## contrast estimate SE df t.ratio p.value
## strep.sp.IMI.cow1 - strep.sp.IMI.cow0 -0.635 1.32 328 -0.482 0.6304
##
## postpartum_sample = 2:
## contrast estimate SE df t.ratio p.value
## strep.sp.IMI.cow1 - strep.sp.IMI.cow0 -2.615 1.30 328 -2.004 0.0458
##
## postpartum_sample = 3:
## contrast estimate SE df t.ratio p.value
## strep.sp.IMI.cow1 - strep.sp.IMI.cow0 -1.346 1.34 328 -1.003 0.3167
##
## postpartum_sample = 4:
## contrast estimate SE df t.ratio p.value
## strep.sp.IMI.cow1 - strep.sp.IMI.cow0 -2.804 1.37 328 -2.047 0.0415
##
## postpartum_sample = 5:
## contrast estimate SE df t.ratio p.value
## strep.sp.IMI.cow1 - strep.sp.IMI.cow0 0.291 1.40 328 0.208 0.8357
##
## postpartum_sample = 6:
## contrast estimate SE df t.ratio p.value
## strep.sp.IMI.cow1 - strep.sp.IMI.cow0 -0.323 1.41 328 -0.229 0.8187
##
## Degrees-of-freedom method: containmentmodel_9 <- lme(fixed =milk_kg ~ strep.sp.IMI.cow*postpartum_sample+calving.nas+calving.sau,random=~1|Site/CowId,correlation= corAR1(form = ~ps_sample),data=df.per.cow_dhi,na.action = na.omit)
emmeans_9 <- emmeans(model_9,revpairwise~strep.sp.IMI.cow|postpartum_sample)
emmeans_9## $emmeans
## postpartum_sample = 1:
## strep.sp.IMI.cow emmean SE df lower.CL upper.CL
## 0 23.5 3.19 3 13.39 33.7
## 1 22.9 3.40 3 12.06 33.7
##
## postpartum_sample = 2:
## strep.sp.IMI.cow emmean SE df lower.CL upper.CL
## 0 25.8 3.19 3 15.63 35.9
## 1 23.2 3.40 3 12.34 34.0
##
## postpartum_sample = 3:
## strep.sp.IMI.cow emmean SE df lower.CL upper.CL
## 0 24.8 3.19 3 14.64 34.9
## 1 23.4 3.41 3 12.58 34.3
##
## postpartum_sample = 4:
## strep.sp.IMI.cow emmean SE df lower.CL upper.CL
## 0 24.0 3.19 3 13.78 34.1
## 1 21.1 3.42 3 10.25 32.0
##
## postpartum_sample = 5:
## strep.sp.IMI.cow emmean SE df lower.CL upper.CL
## 0 23.0 3.19 3 12.88 33.2
## 1 23.3 3.44 3 12.39 34.3
##
## postpartum_sample = 6:
## strep.sp.IMI.cow emmean SE df lower.CL upper.CL
## 0 21.0 3.19 3 10.81 31.1
## 1 20.6 3.44 3 9.71 31.6
##
## Results are averaged over the levels of: calving.nas, calving.sau
## Degrees-of-freedom method: containment
## Confidence level used: 0.95
##
## $contrasts
## postpartum_sample = 1:
## contrast estimate SE df t.ratio p.value
## strep.sp.IMI.cow1 - strep.sp.IMI.cow0 -0.641 1.32 326 -0.485 0.6278
##
## postpartum_sample = 2:
## contrast estimate SE df t.ratio p.value
## strep.sp.IMI.cow1 - strep.sp.IMI.cow0 -2.621 1.31 326 -2.005 0.0458
##
## postpartum_sample = 3:
## contrast estimate SE df t.ratio p.value
## strep.sp.IMI.cow1 - strep.sp.IMI.cow0 -1.352 1.34 326 -1.006 0.3153
##
## postpartum_sample = 4:
## contrast estimate SE df t.ratio p.value
## strep.sp.IMI.cow1 - strep.sp.IMI.cow0 -2.811 1.37 326 -2.048 0.0414
##
## postpartum_sample = 5:
## contrast estimate SE df t.ratio p.value
## strep.sp.IMI.cow1 - strep.sp.IMI.cow0 0.286 1.40 326 0.204 0.8387
##
## postpartum_sample = 6:
## contrast estimate SE df t.ratio p.value
## strep.sp.IMI.cow1 - strep.sp.IMI.cow0 -0.329 1.41 326 -0.233 0.8156
##
## Results are averaged over the levels of: calving.nas, calving.sau
## Degrees-of-freedom method: containmentconfint(emmeans_9)## $emmeans
## postpartum_sample = 1:
## strep.sp.IMI.cow emmean SE df lower.CL upper.CL
## 0 23.5 3.19 3 13.39 33.7
## 1 22.9 3.40 3 12.06 33.7
##
## postpartum_sample = 2:
## strep.sp.IMI.cow emmean SE df lower.CL upper.CL
## 0 25.8 3.19 3 15.63 35.9
## 1 23.2 3.40 3 12.34 34.0
##
## postpartum_sample = 3:
## strep.sp.IMI.cow emmean SE df lower.CL upper.CL
## 0 24.8 3.19 3 14.64 34.9
## 1 23.4 3.41 3 12.58 34.3
##
## postpartum_sample = 4:
## strep.sp.IMI.cow emmean SE df lower.CL upper.CL
## 0 24.0 3.19 3 13.78 34.1
## 1 21.1 3.42 3 10.25 32.0
##
## postpartum_sample = 5:
## strep.sp.IMI.cow emmean SE df lower.CL upper.CL
## 0 23.0 3.19 3 12.88 33.2
## 1 23.3 3.44 3 12.39 34.3
##
## postpartum_sample = 6:
## strep.sp.IMI.cow emmean SE df lower.CL upper.CL
## 0 21.0 3.19 3 10.81 31.1
## 1 20.6 3.44 3 9.71 31.6
##
## Results are averaged over the levels of: calving.nas, calving.sau
## Degrees-of-freedom method: containment
## Confidence level used: 0.95
##
## $contrasts
## postpartum_sample = 1:
## contrast estimate SE df lower.CL upper.CL
## strep.sp.IMI.cow1 - strep.sp.IMI.cow0 -0.641 1.32 326 -3.24 1.957
##
## postpartum_sample = 2:
## contrast estimate SE df lower.CL upper.CL
## strep.sp.IMI.cow1 - strep.sp.IMI.cow0 -2.621 1.31 326 -5.19 -0.049
##
## postpartum_sample = 3:
## contrast estimate SE df lower.CL upper.CL
## strep.sp.IMI.cow1 - strep.sp.IMI.cow0 -1.352 1.34 326 -4.00 1.293
##
## postpartum_sample = 4:
## contrast estimate SE df lower.CL upper.CL
## strep.sp.IMI.cow1 - strep.sp.IMI.cow0 -2.811 1.37 326 -5.51 -0.111
##
## postpartum_sample = 5:
## contrast estimate SE df lower.CL upper.CL
## strep.sp.IMI.cow1 - strep.sp.IMI.cow0 0.286 1.40 326 -2.48 3.050
##
## postpartum_sample = 6:
## contrast estimate SE df lower.CL upper.CL
## strep.sp.IMI.cow1 - strep.sp.IMI.cow0 -0.329 1.41 326 -3.10 2.444
##
## Results are averaged over the levels of: calving.nas, calving.sau
## Degrees-of-freedom method: containment
## Confidence level used: 0.95adj_means_9 <- as.data.frame(emmeans_9$emmeans)
adj_means_9## postpartum_sample = 1:
## strep.sp.IMI.cow emmean SE df lower.CL upper.CL
## 0 23.53571 3.187823 3 13.390631 33.68078
## 1 22.89479 3.404624 3 12.059763 33.72983
##
## postpartum_sample = 2:
## strep.sp.IMI.cow emmean SE df lower.CL upper.CL
## 0 25.77887 3.188014 3 15.633185 35.92455
## 1 23.15820 3.399783 3 12.338570 33.97782
##
## postpartum_sample = 3:
## strep.sp.IMI.cow emmean SE df lower.CL upper.CL
## 0 24.79177 3.189373 3 14.641756 34.94177
## 1 23.43943 3.413121 3 12.577353 34.30150
##
## postpartum_sample = 4:
## strep.sp.IMI.cow emmean SE df lower.CL upper.CL
## 0 23.95079 3.194610 3 13.784117 34.11747
## 1 21.13975 3.421827 3 10.249970 32.02953
##
## postpartum_sample = 5:
## strep.sp.IMI.cow emmean SE df lower.CL upper.CL
## 0 23.03544 3.191821 3 12.877638 33.19324
## 1 23.32163 3.435992 3 12.386766 34.25648
##
## postpartum_sample = 6:
## strep.sp.IMI.cow emmean SE df lower.CL upper.CL
## 0 20.97381 3.192976 3 10.812339 31.13529
## 1 20.64470 3.437415 3 9.705308 31.58408
##
## Results are averaged over the levels of: calving.nas, calving.sau
## Degrees-of-freedom method: containment
## Confidence level used: 0.95contrasts_9 <- as.data.frame(emmeans_9$contrasts) %>% select (postpartum_sample,p.value)
adj_means_9 <- adj_means_9 %>% left_join(contrasts_9, by="postpartum_sample")
adj_means_9 <- adj_means_9 %>% group_by(postpartum_sample) %>% mutate(max_milk = max(upper.CL)) %>% ungroup()
adj_means_9 <- adj_means_9 %>% mutate(p.value = ifelse(max_milk == upper.CL, p.value, 1))
strep_sp_title <- expression(paste("Unspeciated ",italic("Streptococcus ")))
milk_9 <- ggplot(adj_means_9, aes(x=postpartum_sample, y=emmean, group=strep.sp.IMI.cow, color=strep.sp.IMI.cow)) +
geom_line() +
geom_point()+
geom_errorbar(aes(ymin=emmean-SE, ymax=emmean+SE), width=.2,
position=position_dodge(0.05)) +
ylim(0,32) +
labs(x="Monthly postpartum test",y="Milk production (Kg/day)",title=strep_sp_title,color="IMI") +
theme_classic() +
theme(plot.title=element_text(face="italic")) +
scale_color_viridis(discrete = TRUE) +
scale_color_manual(labels = c("No", "Yes"),
values = c("#009999", "#990000")) +
geom_text(aes(label = ifelse(p.value < 0.05, "*", "")),
vjust = -3, size = 5, color = "black")## Scale for colour is already present.
## Adding another scale for colour, which will replace the existing scale.model_9_adj_overall <- lme(fixed =milk_kg~ strep.sp.IMI.cow+calving.sau+calving.nas,random=~1|Site/CowId,correlation= corAR1(form = ~ps_sample),data=df.per.cow_dhi,na.action = na.omit)
Confint(model_9_adj_overall)## Estimate 2.5 % 97.5 %
## (Intercept) 23.8315558 17.638362 30.0247496
## strep.sp.IMI.cow1 -1.2385084 -3.029250 0.5522331
## calving.sau1 -0.1286207 -1.560117 1.3028755
## calving.nas1 -0.1416688 -1.340876 1.0575382# ----Streptococcus like organisms ----
model_10 <- lme(fixed =milk_kg~ strep.like.IMI.cow*postpartum_sample,random=~1|Site/CowId,correlation= corAR1(form = ~ps_sample),data=df.per.cow_dhi,na.action = na.omit)
emmeans(model_10,revpairwise~strep.like.IMI.cow|postpartum_sample)## $emmeans
## postpartum_sample = 1:
## strep.like.IMI.cow emmean SE df lower.CL upper.CL
## 0 23.6 3.20 3 13.4 33.8
## 1 23.3 3.30 3 12.8 33.8
##
## postpartum_sample = 2:
## strep.like.IMI.cow emmean SE df lower.CL upper.CL
## 0 25.6 3.20 3 15.4 35.8
## 1 25.4 3.31 3 14.9 36.0
##
## postpartum_sample = 3:
## strep.like.IMI.cow emmean SE df lower.CL upper.CL
## 0 24.7 3.20 3 14.5 34.9
## 1 24.8 3.33 3 14.2 35.4
##
## postpartum_sample = 4:
## strep.like.IMI.cow emmean SE df lower.CL upper.CL
## 0 23.7 3.21 3 13.5 33.9
## 1 23.7 3.34 3 13.1 34.4
##
## postpartum_sample = 5:
## strep.like.IMI.cow emmean SE df lower.CL upper.CL
## 0 23.3 3.20 3 13.1 33.5
## 1 22.4 3.34 3 11.8 33.0
##
## postpartum_sample = 6:
## strep.like.IMI.cow emmean SE df lower.CL upper.CL
## 0 21.1 3.21 3 10.9 31.3
## 1 20.8 3.35 3 10.1 31.5
##
## Degrees-of-freedom method: containment
## Confidence level used: 0.95
##
## $contrasts
## postpartum_sample = 1:
## contrast estimate SE df t.ratio p.value
## strep.like.IMI.cow1 - strep.like.IMI.cow0 -0.30083 1.01 328 -0.298 0.7660
##
## postpartum_sample = 2:
## contrast estimate SE df t.ratio p.value
## strep.like.IMI.cow1 - strep.like.IMI.cow0 -0.19489 1.04 328 -0.188 0.8511
##
## postpartum_sample = 3:
## contrast estimate SE df t.ratio p.value
## strep.like.IMI.cow1 - strep.like.IMI.cow0 0.09261 1.09 328 0.085 0.9326
##
## postpartum_sample = 4:
## contrast estimate SE df t.ratio p.value
## strep.like.IMI.cow1 - strep.like.IMI.cow0 0.00183 1.15 328 0.002 0.9987
##
## postpartum_sample = 5:
## contrast estimate SE df t.ratio p.value
## strep.like.IMI.cow1 - strep.like.IMI.cow0 -0.86566 1.14 328 -0.760 0.4478
##
## postpartum_sample = 6:
## contrast estimate SE df t.ratio p.value
## strep.like.IMI.cow1 - strep.like.IMI.cow0 -0.27872 1.18 328 -0.236 0.8132
##
## Degrees-of-freedom method: containmentmodel_10 <- lme(fixed =milk_kg~ strep.like.IMI.cow*postpartum_sample+calving.nas+calving.sau,random=~1|Site/CowId,correlation= corAR1(form = ~ps_sample),data=df.per.cow_dhi,na.action = na.omit)
emmeans_10 <- emmeans(model_10,revpairwise~strep.like.IMI.cow|postpartum_sample)
emmeans_10## $emmeans
## postpartum_sample = 1:
## strep.like.IMI.cow emmean SE df lower.CL upper.CL
## 0 23.5 3.20 3 13.34 33.7
## 1 23.2 3.31 3 12.67 33.7
##
## postpartum_sample = 2:
## strep.like.IMI.cow emmean SE df lower.CL upper.CL
## 0 25.6 3.20 3 15.37 35.8
## 1 25.3 3.32 3 14.78 35.9
##
## postpartum_sample = 3:
## strep.like.IMI.cow emmean SE df lower.CL upper.CL
## 0 24.7 3.20 3 14.45 34.8
## 1 24.7 3.34 3 14.10 35.3
##
## postpartum_sample = 4:
## strep.like.IMI.cow emmean SE df lower.CL upper.CL
## 0 23.6 3.21 3 13.43 33.9
## 1 23.6 3.35 3 12.96 34.3
##
## postpartum_sample = 5:
## strep.like.IMI.cow emmean SE df lower.CL upper.CL
## 0 23.2 3.21 3 13.00 33.4
## 1 22.3 3.35 3 11.66 33.0
##
## postpartum_sample = 6:
## strep.like.IMI.cow emmean SE df lower.CL upper.CL
## 0 21.0 3.21 3 10.78 31.2
## 1 20.7 3.36 3 9.99 31.4
##
## Results are averaged over the levels of: calving.nas, calving.sau
## Degrees-of-freedom method: containment
## Confidence level used: 0.95
##
## $contrasts
## postpartum_sample = 1:
## contrast estimate SE df t.ratio p.value
## strep.like.IMI.cow1 - strep.like.IMI.cow0 -0.3252 1.02 326 -0.320 0.7489
##
## postpartum_sample = 2:
## contrast estimate SE df t.ratio p.value
## strep.like.IMI.cow1 - strep.like.IMI.cow0 -0.2197 1.04 326 -0.211 0.8333
##
## postpartum_sample = 3:
## contrast estimate SE df t.ratio p.value
## strep.like.IMI.cow1 - strep.like.IMI.cow0 0.0696 1.10 326 0.063 0.9495
##
## postpartum_sample = 4:
## contrast estimate SE df t.ratio p.value
## strep.like.IMI.cow1 - strep.like.IMI.cow0 -0.0212 1.15 326 -0.018 0.9854
##
## postpartum_sample = 5:
## contrast estimate SE df t.ratio p.value
## strep.like.IMI.cow1 - strep.like.IMI.cow0 -0.8884 1.14 326 -0.777 0.4378
##
## postpartum_sample = 6:
## contrast estimate SE df t.ratio p.value
## strep.like.IMI.cow1 - strep.like.IMI.cow0 -0.3017 1.18 326 -0.255 0.7989
##
## Results are averaged over the levels of: calving.nas, calving.sau
## Degrees-of-freedom method: containmentconfint(emmeans_10)## $emmeans
## postpartum_sample = 1:
## strep.like.IMI.cow emmean SE df lower.CL upper.CL
## 0 23.5 3.20 3 13.34 33.7
## 1 23.2 3.31 3 12.67 33.7
##
## postpartum_sample = 2:
## strep.like.IMI.cow emmean SE df lower.CL upper.CL
## 0 25.6 3.20 3 15.37 35.8
## 1 25.3 3.32 3 14.78 35.9
##
## postpartum_sample = 3:
## strep.like.IMI.cow emmean SE df lower.CL upper.CL
## 0 24.7 3.20 3 14.45 34.8
## 1 24.7 3.34 3 14.10 35.3
##
## postpartum_sample = 4:
## strep.like.IMI.cow emmean SE df lower.CL upper.CL
## 0 23.6 3.21 3 13.43 33.9
## 1 23.6 3.35 3 12.96 34.3
##
## postpartum_sample = 5:
## strep.like.IMI.cow emmean SE df lower.CL upper.CL
## 0 23.2 3.21 3 13.00 33.4
## 1 22.3 3.35 3 11.66 33.0
##
## postpartum_sample = 6:
## strep.like.IMI.cow emmean SE df lower.CL upper.CL
## 0 21.0 3.21 3 10.78 31.2
## 1 20.7 3.36 3 9.99 31.4
##
## Results are averaged over the levels of: calving.nas, calving.sau
## Degrees-of-freedom method: containment
## Confidence level used: 0.95
##
## $contrasts
## postpartum_sample = 1:
## contrast estimate SE df lower.CL upper.CL
## strep.like.IMI.cow1 - strep.like.IMI.cow0 -0.3252 1.02 326 -2.32 1.67
##
## postpartum_sample = 2:
## contrast estimate SE df lower.CL upper.CL
## strep.like.IMI.cow1 - strep.like.IMI.cow0 -0.2197 1.04 326 -2.27 1.83
##
## postpartum_sample = 3:
## contrast estimate SE df lower.CL upper.CL
## strep.like.IMI.cow1 - strep.like.IMI.cow0 0.0696 1.10 326 -2.09 2.23
##
## postpartum_sample = 4:
## contrast estimate SE df lower.CL upper.CL
## strep.like.IMI.cow1 - strep.like.IMI.cow0 -0.0212 1.15 326 -2.29 2.25
##
## postpartum_sample = 5:
## contrast estimate SE df lower.CL upper.CL
## strep.like.IMI.cow1 - strep.like.IMI.cow0 -0.8884 1.14 326 -3.14 1.36
##
## postpartum_sample = 6:
## contrast estimate SE df lower.CL upper.CL
## strep.like.IMI.cow1 - strep.like.IMI.cow0 -0.3017 1.18 326 -2.63 2.03
##
## Results are averaged over the levels of: calving.nas, calving.sau
## Degrees-of-freedom method: containment
## Confidence level used: 0.95adj_means_10 <- as.data.frame(emmeans_10$emmeans)
adj_means_10## postpartum_sample = 1:
## strep.like.IMI.cow emmean SE df lower.CL upper.CL
## 0 23.53111 3.202881 3 13.338115 33.72411
## 1 23.20590 3.311423 3 12.667471 33.74432
##
## postpartum_sample = 2:
## strep.like.IMI.cow emmean SE df lower.CL upper.CL
## 0 25.56298 3.202835 3 15.370126 35.75583
## 1 25.34328 3.319474 3 14.779234 35.90733
##
## postpartum_sample = 3:
## strep.like.IMI.cow emmean SE df lower.CL upper.CL
## 0 24.65116 3.204072 3 14.454378 34.84795
## 1 24.72076 3.335896 3 14.104447 35.33707
##
## postpartum_sample = 4:
## strep.like.IMI.cow emmean SE df lower.CL upper.CL
## 0 23.64900 3.209689 3 13.434343 33.86367
## 1 23.62783 3.351032 3 12.963346 34.29230
##
## postpartum_sample = 5:
## strep.like.IMI.cow emmean SE df lower.CL upper.CL
## 0 23.20356 3.206786 3 12.998139 33.40899
## 1 22.31520 3.348774 3 11.657905 32.97249
##
## postpartum_sample = 6:
## strep.like.IMI.cow emmean SE df lower.CL upper.CL
## 0 20.99198 3.207710 3 10.783615 31.20034
## 1 20.69024 3.361775 3 9.991574 31.38891
##
## Results are averaged over the levels of: calving.nas, calving.sau
## Degrees-of-freedom method: containment
## Confidence level used: 0.95contrasts_10 <- as.data.frame(emmeans_10$contrasts) %>% select (postpartum_sample,p.value)
adj_means_10 <- adj_means_10 %>% left_join(contrasts_10, by="postpartum_sample")
adj_means_10 <- adj_means_10 %>% group_by(postpartum_sample) %>% mutate(max_milk = max(upper.CL)) %>% ungroup()
adj_means_10 <- adj_means_10 %>% mutate(p.value = ifelse(max_milk == upper.CL, p.value, 1))
strep_like_title <- expression(paste(italic("Streptococcus"),"-like organisms"))
milk_10 <- ggplot(adj_means_10, aes(x=postpartum_sample, y=emmean, group=strep.like.IMI.cow, color=strep.like.IMI.cow)) +
geom_line() +
geom_point()+
geom_errorbar(aes(ymin=emmean-SE, ymax=emmean+SE), width=.2,
position=position_dodge(0.05)) +
ylim(0,32) +
labs(x="Monthly postpartum test",y="Milk production (Kg/day)",title=strep_like_title,color="IMI") +
theme_classic() +
theme(plot.title=element_text(face="italic")) +
scale_color_viridis(discrete = TRUE) +
scale_color_manual(labels = c("No", "Yes"),
values = c("#009999", "#990000")) +
geom_text(aes(label = ifelse(p.value < 0.05, "*", "")),
vjust = -3, size = 5, color = "black")## Scale for colour is already present.
## Adding another scale for colour, which will replace the existing scale.model_10_adj_overall <- lme(fixed =milk_kg~ strep.like.IMI.cow+calving.sau+calving.nas,random=~1|Site/CowId,correlation= corAR1(form = ~ps_sample),data=df.per.cow_dhi,na.action = na.omit)
Confint(model_10_adj_overall)## Estimate 2.5 % 97.5 %
## (Intercept) 23.7525373 17.533913 29.971162
## strep.like.IMI.cow1 -0.1848530 -1.663962 1.294256
## calving.sau1 -0.1528709 -1.593852 1.288110
## calving.nas1 -0.1249329 -1.332404 1.082539# ---- Enterococcus genus ----
model_11 <- lme(fixed =milk_kg~ enterococcus.spp.IMI.cow*postpartum_sample,random=~1|Site/CowId,correlation= corAR1(form = ~ps_sample),data=df.per.cow_dhi,na.action = na.omit)
emmeans(model_11,revpairwise~enterococcus.spp.IMI.cow|postpartum_sample)## $emmeans
## postpartum_sample = 1:
## enterococcus.spp.IMI.cow emmean SE df lower.CL upper.CL
## 0 23.7 3.22 3 13.47 33.9
## 1 22.3 3.39 3 11.53 33.1
##
## postpartum_sample = 2:
## enterococcus.spp.IMI.cow emmean SE df lower.CL upper.CL
## 0 25.6 3.22 3 15.38 35.9
## 1 25.3 3.40 3 14.51 36.2
##
## postpartum_sample = 3:
## enterococcus.spp.IMI.cow emmean SE df lower.CL upper.CL
## 0 24.7 3.22 3 14.46 34.9
## 1 24.9 3.43 3 14.01 35.9
##
## postpartum_sample = 4:
## enterococcus.spp.IMI.cow emmean SE df lower.CL upper.CL
## 0 23.7 3.22 3 13.42 33.9
## 1 23.9 3.44 3 12.97 34.9
##
## postpartum_sample = 5:
## enterococcus.spp.IMI.cow emmean SE df lower.CL upper.CL
## 0 23.2 3.22 3 12.99 33.5
## 1 22.2 3.45 3 11.22 33.2
##
## postpartum_sample = 6:
## enterococcus.spp.IMI.cow emmean SE df lower.CL upper.CL
## 0 21.1 3.22 3 10.81 31.3
## 1 20.6 3.46 3 9.59 31.6
##
## Degrees-of-freedom method: containment
## Confidence level used: 0.95
##
## $contrasts
## postpartum_sample = 1:
## contrast estimate SE df
## enterococcus.spp.IMI.cow1 - enterococcus.spp.IMI.cow0 -1.374 1.22 328
## t.ratio p.value
## -1.125 0.2615
##
## postpartum_sample = 2:
## contrast estimate SE df
## enterococcus.spp.IMI.cow1 - enterococcus.spp.IMI.cow0 -0.279 1.25 328
## t.ratio p.value
## -0.223 0.8240
##
## postpartum_sample = 3:
## contrast estimate SE df
## enterococcus.spp.IMI.cow1 - enterococcus.spp.IMI.cow0 0.234 1.33 328
## t.ratio p.value
## 0.175 0.8611
##
## postpartum_sample = 4:
## contrast estimate SE df
## enterococcus.spp.IMI.cow1 - enterococcus.spp.IMI.cow0 0.240 1.36 328
## t.ratio p.value
## 0.176 0.8600
##
## postpartum_sample = 5:
## contrast estimate SE df
## enterococcus.spp.IMI.cow1 - enterococcus.spp.IMI.cow0 -1.042 1.38 328
## t.ratio p.value
## -0.757 0.4495
##
## postpartum_sample = 6:
## contrast estimate SE df
## enterococcus.spp.IMI.cow1 - enterococcus.spp.IMI.cow0 -0.456 1.41 328
## t.ratio p.value
## -0.323 0.7472
##
## Degrees-of-freedom method: containmentmodel_11 <- lme(fixed =milk_kg~ enterococcus.spp.IMI.cow*postpartum_sample+calving.nas+calving.sau,random=~1|Site/CowId,correlation= corAR1(form = ~ps_sample),data=df.per.cow_dhi,na.action = na.omit)
emmeans_11 <- emmeans(model_11,revpairwise~enterococcus.spp.IMI.cow|postpartum_sample)
emmeans_11## $emmeans
## postpartum_sample = 1:
## enterococcus.spp.IMI.cow emmean SE df lower.CL upper.CL
## 0 23.6 3.22 3 13.39 33.9
## 1 22.2 3.40 3 11.41 33.1
##
## postpartum_sample = 2:
## enterococcus.spp.IMI.cow emmean SE df lower.CL upper.CL
## 0 25.5 3.22 3 15.30 35.8
## 1 25.2 3.41 3 14.39 36.1
##
## postpartum_sample = 3:
## enterococcus.spp.IMI.cow emmean SE df lower.CL upper.CL
## 0 24.6 3.22 3 14.38 34.9
## 1 24.8 3.44 3 13.89 35.8
##
## postpartum_sample = 4:
## enterococcus.spp.IMI.cow emmean SE df lower.CL upper.CL
## 0 23.6 3.23 3 13.34 33.9
## 1 23.8 3.45 3 12.85 34.8
##
## postpartum_sample = 5:
## enterococcus.spp.IMI.cow emmean SE df lower.CL upper.CL
## 0 23.2 3.22 3 12.91 33.4
## 1 22.1 3.46 3 11.10 33.1
##
## postpartum_sample = 6:
## enterococcus.spp.IMI.cow emmean SE df lower.CL upper.CL
## 0 21.0 3.22 3 10.72 31.2
## 1 20.5 3.47 3 9.46 31.6
##
## Results are averaged over the levels of: calving.nas, calving.sau
## Degrees-of-freedom method: containment
## Confidence level used: 0.95
##
## $contrasts
## postpartum_sample = 1:
## contrast estimate SE df
## enterococcus.spp.IMI.cow1 - enterococcus.spp.IMI.cow0 -1.396 1.23 326
## t.ratio p.value
## -1.138 0.2559
##
## postpartum_sample = 2:
## contrast estimate SE df
## enterococcus.spp.IMI.cow1 - enterococcus.spp.IMI.cow0 -0.301 1.26 326
## t.ratio p.value
## -0.239 0.8112
##
## postpartum_sample = 3:
## contrast estimate SE df
## enterococcus.spp.IMI.cow1 - enterococcus.spp.IMI.cow0 0.212 1.34 326
## t.ratio p.value
## 0.158 0.8743
##
## postpartum_sample = 4:
## contrast estimate SE df
## enterococcus.spp.IMI.cow1 - enterococcus.spp.IMI.cow0 0.220 1.37 326
## t.ratio p.value
## 0.161 0.8724
##
## postpartum_sample = 5:
## contrast estimate SE df
## enterococcus.spp.IMI.cow1 - enterococcus.spp.IMI.cow0 -1.062 1.38 326
## t.ratio p.value
## -0.769 0.4422
##
## postpartum_sample = 6:
## contrast estimate SE df
## enterococcus.spp.IMI.cow1 - enterococcus.spp.IMI.cow0 -0.477 1.42 326
## t.ratio p.value
## -0.337 0.7367
##
## Results are averaged over the levels of: calving.nas, calving.sau
## Degrees-of-freedom method: containmentconfint(emmeans_11)## $emmeans
## postpartum_sample = 1:
## enterococcus.spp.IMI.cow emmean SE df lower.CL upper.CL
## 0 23.6 3.22 3 13.39 33.9
## 1 22.2 3.40 3 11.41 33.1
##
## postpartum_sample = 2:
## enterococcus.spp.IMI.cow emmean SE df lower.CL upper.CL
## 0 25.5 3.22 3 15.30 35.8
## 1 25.2 3.41 3 14.39 36.1
##
## postpartum_sample = 3:
## enterococcus.spp.IMI.cow emmean SE df lower.CL upper.CL
## 0 24.6 3.22 3 14.38 34.9
## 1 24.8 3.44 3 13.89 35.8
##
## postpartum_sample = 4:
## enterococcus.spp.IMI.cow emmean SE df lower.CL upper.CL
## 0 23.6 3.23 3 13.34 33.9
## 1 23.8 3.45 3 12.85 34.8
##
## postpartum_sample = 5:
## enterococcus.spp.IMI.cow emmean SE df lower.CL upper.CL
## 0 23.2 3.22 3 12.91 33.4
## 1 22.1 3.46 3 11.10 33.1
##
## postpartum_sample = 6:
## enterococcus.spp.IMI.cow emmean SE df lower.CL upper.CL
## 0 21.0 3.22 3 10.72 31.2
## 1 20.5 3.47 3 9.46 31.6
##
## Results are averaged over the levels of: calving.nas, calving.sau
## Degrees-of-freedom method: containment
## Confidence level used: 0.95
##
## $contrasts
## postpartum_sample = 1:
## contrast estimate SE df
## enterococcus.spp.IMI.cow1 - enterococcus.spp.IMI.cow0 -1.396 1.23 326
## lower.CL upper.CL
## -3.81 1.02
##
## postpartum_sample = 2:
## contrast estimate SE df
## enterococcus.spp.IMI.cow1 - enterococcus.spp.IMI.cow0 -0.301 1.26 326
## lower.CL upper.CL
## -2.78 2.17
##
## postpartum_sample = 3:
## contrast estimate SE df
## enterococcus.spp.IMI.cow1 - enterococcus.spp.IMI.cow0 0.212 1.34 326
## lower.CL upper.CL
## -2.42 2.84
##
## postpartum_sample = 4:
## contrast estimate SE df
## enterococcus.spp.IMI.cow1 - enterococcus.spp.IMI.cow0 0.220 1.37 326
## lower.CL upper.CL
## -2.47 2.91
##
## postpartum_sample = 5:
## contrast estimate SE df
## enterococcus.spp.IMI.cow1 - enterococcus.spp.IMI.cow0 -1.062 1.38 326
## lower.CL upper.CL
## -3.78 1.65
##
## postpartum_sample = 6:
## contrast estimate SE df
## enterococcus.spp.IMI.cow1 - enterococcus.spp.IMI.cow0 -0.477 1.42 326
## lower.CL upper.CL
## -3.26 2.31
##
## Results are averaged over the levels of: calving.nas, calving.sau
## Degrees-of-freedom method: containment
## Confidence level used: 0.95adj_means_11 <- as.data.frame(emmeans_11$emmeans)
adj_means_11## postpartum_sample = 1:
## enterococcus.spp.IMI.cow emmean SE df lower.CL upper.CL
## 0 23.63360 3.219617 3 13.387341 33.87986
## 1 22.23748 3.402552 3 11.409042 33.06592
##
## postpartum_sample = 2:
## enterococcus.spp.IMI.cow emmean SE df lower.CL upper.CL
## 0 25.54884 3.219719 3 15.302262 35.79543
## 1 25.24797 3.413364 3 14.385127 36.11082
##
## postpartum_sample = 3:
## enterococcus.spp.IMI.cow emmean SE df lower.CL upper.CL
## 0 24.63362 3.220840 3 14.383474 34.88377
## 1 24.84549 3.442505 3 13.889906 35.80108
##
## postpartum_sample = 4:
## enterococcus.spp.IMI.cow emmean SE df lower.CL upper.CL
## 0 23.60780 3.226201 3 13.340591 33.87501
## 1 23.82751 3.450570 3 12.846256 34.80876
##
## postpartum_sample = 5:
## enterococcus.spp.IMI.cow emmean SE df lower.CL upper.CL
## 0 23.16422 3.223425 3 12.905844 33.42260
## 1 22.10243 3.457187 3 11.100113 33.10474
##
## postpartum_sample = 6:
## enterococcus.spp.IMI.cow emmean SE df lower.CL upper.CL
## 0 20.98585 3.224451 3 10.724203 31.24749
## 1 20.50919 3.471020 3 9.462858 31.55553
##
## Results are averaged over the levels of: calving.nas, calving.sau
## Degrees-of-freedom method: containment
## Confidence level used: 0.95contrasts_11 <- as.data.frame(emmeans_11$contrasts) %>% select (postpartum_sample,p.value)
adj_means_11 <- adj_means_11 %>% left_join(contrasts_11, by="postpartum_sample")
adj_means_11 <- adj_means_11 %>% group_by(postpartum_sample) %>% mutate(max_milk = max(upper.CL)) %>% ungroup()
adj_means_11 <- adj_means_11 %>% mutate(p.value = ifelse(max_milk == upper.CL, p.value, 1))
enterococcus_genus_title <- expression(paste(italic("Enterococcus "),"spp."))
adj_means_11$enterococcus.spp.IMI.cow <- as.factor(adj_means_11$enterococcus.spp.IMI.cow)
milk_11 <- ggplot(adj_means_11, aes(x=postpartum_sample, y=emmean, group=enterococcus.spp.IMI.cow, color=enterococcus.spp.IMI.cow)) +
geom_line() +
geom_point()+
geom_errorbar(aes(ymin=emmean-SE, ymax=emmean+SE), width=.2,
position=position_dodge(0.05)) +
ylim(0,32) +
labs(x="Monthly postpartum test",y="Milk production (Kg/day)",title=enterococcus_genus_title,color="IMI") +
theme_classic() +
theme(plot.title=element_text(face="italic")) +
scale_color_viridis(discrete = TRUE) +
scale_color_manual(labels = c("No", "Yes"),
values = c("#009999", "#990000")) +
geom_text(aes(label = ifelse(p.value < 0.05, "*", "")),
vjust = -3, size = 5, color = "black")## Scale for colour is already present.
## Adding another scale for colour, which will replace the existing scale.model_11_adj_overall <- lme(fixed =milk_kg~ strep.like.IMI.cow+calving.sau+calving.nas,random=~1|Site/CowId,correlation= corAR1(form = ~ps_sample),data=df.per.cow_dhi,na.action = na.omit)
Confint(model_11_adj_overall)## Estimate 2.5 % 97.5 %
## (Intercept) 23.7525373 17.533913 29.971162
## strep.like.IMI.cow1 -0.1848530 -1.663962 1.294256
## calving.sau1 -0.1528709 -1.593852 1.288110
## calving.nas1 -0.1249329 -1.332404 1.082539# --- Gram negative bacteria ----
model_12 <- lme(fixed =milk_kg~ gram.neg.IMI.cow*postpartum_sample,random=~1|Site/CowId,correlation= corAR1(form = ~ps_sample),data=df.per.cow_dhi,na.action = na.omit)
emmeans(model_12,revpairwise~gram.neg.IMI.cow|postpartum_sample)## $emmeans
## postpartum_sample = 1:
## gram.neg.IMI.cow emmean SE df lower.CL upper.CL
## 0 23.7 3.20 3 13.5 33.9
## 1 22.0 3.42 3 11.1 32.9
##
## postpartum_sample = 2:
## gram.neg.IMI.cow emmean SE df lower.CL upper.CL
## 0 25.6 3.20 3 15.4 35.8
## 1 25.4 3.42 3 14.5 36.3
##
## postpartum_sample = 3:
## gram.neg.IMI.cow emmean SE df lower.CL upper.CL
## 0 24.7 3.20 3 14.5 34.9
## 1 25.2 3.43 3 14.3 36.1
##
## postpartum_sample = 4:
## gram.neg.IMI.cow emmean SE df lower.CL upper.CL
## 0 23.7 3.21 3 13.5 33.9
## 1 23.6 3.44 3 12.7 34.6
##
## postpartum_sample = 5:
## gram.neg.IMI.cow emmean SE df lower.CL upper.CL
## 0 23.1 3.20 3 12.9 33.3
## 1 23.7 3.45 3 12.7 34.7
##
## postpartum_sample = 6:
## gram.neg.IMI.cow emmean SE df lower.CL upper.CL
## 0 20.9 3.20 3 10.7 31.1
## 1 22.5 3.48 3 11.4 33.6
##
## Degrees-of-freedom method: containment
## Confidence level used: 0.95
##
## $contrasts
## postpartum_sample = 1:
## contrast estimate SE df t.ratio p.value
## gram.neg.IMI.cow1 - gram.neg.IMI.cow0 -1.683 1.32 328 -1.270 0.2049
##
## postpartum_sample = 2:
## contrast estimate SE df t.ratio p.value
## gram.neg.IMI.cow1 - gram.neg.IMI.cow0 -0.241 1.33 328 -0.181 0.8562
##
## postpartum_sample = 3:
## contrast estimate SE df t.ratio p.value
## gram.neg.IMI.cow1 - gram.neg.IMI.cow0 0.510 1.36 328 0.374 0.7085
##
## postpartum_sample = 4:
## contrast estimate SE df t.ratio p.value
## gram.neg.IMI.cow1 - gram.neg.IMI.cow0 -0.123 1.39 328 -0.089 0.9294
##
## postpartum_sample = 5:
## contrast estimate SE df t.ratio p.value
## gram.neg.IMI.cow1 - gram.neg.IMI.cow0 0.639 1.42 328 0.449 0.6541
##
## postpartum_sample = 6:
## contrast estimate SE df t.ratio p.value
## gram.neg.IMI.cow1 - gram.neg.IMI.cow0 1.638 1.49 328 1.100 0.2722
##
## Degrees-of-freedom method: containmentmodel_12 <- lme(fixed =milk_kg~ gram.neg.IMI.cow*postpartum_sample+calving.nas+calving.sau.sslo,random=~1|Site/CowId,correlation= corAR1(form = ~ps_sample),data=df.per.cow_dhi,na.action = na.omit)
emmeans_12 <- emmeans(model_12,revpairwise~gram.neg.IMI.cow|postpartum_sample)
emmeans_12## $emmeans
## postpartum_sample = 1:
## gram.neg.IMI.cow emmean SE df lower.CL upper.CL
## 0 23.3 3.09 3 13.5 33.2
## 1 21.7 3.32 3 11.1 32.2
##
## postpartum_sample = 2:
## gram.neg.IMI.cow emmean SE df lower.CL upper.CL
## 0 25.2 3.09 3 15.4 35.1
## 1 25.0 3.32 3 14.5 35.6
##
## postpartum_sample = 3:
## gram.neg.IMI.cow emmean SE df lower.CL upper.CL
## 0 24.3 3.10 3 14.5 34.2
## 1 24.8 3.33 3 14.2 35.4
##
## postpartum_sample = 4:
## gram.neg.IMI.cow emmean SE df lower.CL upper.CL
## 0 23.4 3.10 3 13.5 33.2
## 1 23.2 3.34 3 12.6 33.8
##
## postpartum_sample = 5:
## gram.neg.IMI.cow emmean SE df lower.CL upper.CL
## 0 22.7 3.10 3 12.8 32.6
## 1 23.3 3.36 3 12.6 34.0
##
## postpartum_sample = 6:
## gram.neg.IMI.cow emmean SE df lower.CL upper.CL
## 0 20.5 3.10 3 10.6 30.4
## 1 22.1 3.38 3 11.4 32.9
##
## Results are averaged over the levels of: calving.nas, calving.sau.sslo
## Degrees-of-freedom method: containment
## Confidence level used: 0.95
##
## $contrasts
## postpartum_sample = 1:
## contrast estimate SE df t.ratio p.value
## gram.neg.IMI.cow1 - gram.neg.IMI.cow0 -1.686 1.32 326 -1.277 0.2024
##
## postpartum_sample = 2:
## contrast estimate SE df t.ratio p.value
## gram.neg.IMI.cow1 - gram.neg.IMI.cow0 -0.229 1.32 326 -0.174 0.8622
##
## postpartum_sample = 3:
## contrast estimate SE df t.ratio p.value
## gram.neg.IMI.cow1 - gram.neg.IMI.cow0 0.501 1.36 326 0.369 0.7120
##
## postpartum_sample = 4:
## contrast estimate SE df t.ratio p.value
## gram.neg.IMI.cow1 - gram.neg.IMI.cow0 -0.148 1.39 326 -0.107 0.9148
##
## postpartum_sample = 5:
## contrast estimate SE df t.ratio p.value
## gram.neg.IMI.cow1 - gram.neg.IMI.cow0 0.615 1.42 326 0.434 0.6647
##
## postpartum_sample = 6:
## contrast estimate SE df t.ratio p.value
## gram.neg.IMI.cow1 - gram.neg.IMI.cow0 1.635 1.48 326 1.101 0.2716
##
## Results are averaged over the levels of: calving.nas, calving.sau.sslo
## Degrees-of-freedom method: containmentconfint(emmeans_12)## $emmeans
## postpartum_sample = 1:
## gram.neg.IMI.cow emmean SE df lower.CL upper.CL
## 0 23.3 3.09 3 13.5 33.2
## 1 21.7 3.32 3 11.1 32.2
##
## postpartum_sample = 2:
## gram.neg.IMI.cow emmean SE df lower.CL upper.CL
## 0 25.2 3.09 3 15.4 35.1
## 1 25.0 3.32 3 14.5 35.6
##
## postpartum_sample = 3:
## gram.neg.IMI.cow emmean SE df lower.CL upper.CL
## 0 24.3 3.10 3 14.5 34.2
## 1 24.8 3.33 3 14.2 35.4
##
## postpartum_sample = 4:
## gram.neg.IMI.cow emmean SE df lower.CL upper.CL
## 0 23.4 3.10 3 13.5 33.2
## 1 23.2 3.34 3 12.6 33.8
##
## postpartum_sample = 5:
## gram.neg.IMI.cow emmean SE df lower.CL upper.CL
## 0 22.7 3.10 3 12.8 32.6
## 1 23.3 3.36 3 12.6 34.0
##
## postpartum_sample = 6:
## gram.neg.IMI.cow emmean SE df lower.CL upper.CL
## 0 20.5 3.10 3 10.6 30.4
## 1 22.1 3.38 3 11.4 32.9
##
## Results are averaged over the levels of: calving.nas, calving.sau.sslo
## Degrees-of-freedom method: containment
## Confidence level used: 0.95
##
## $contrasts
## postpartum_sample = 1:
## contrast estimate SE df lower.CL upper.CL
## gram.neg.IMI.cow1 - gram.neg.IMI.cow0 -1.686 1.32 326 -4.28 0.911
##
## postpartum_sample = 2:
## contrast estimate SE df lower.CL upper.CL
## gram.neg.IMI.cow1 - gram.neg.IMI.cow0 -0.229 1.32 326 -2.83 2.370
##
## postpartum_sample = 3:
## contrast estimate SE df lower.CL upper.CL
## gram.neg.IMI.cow1 - gram.neg.IMI.cow0 0.501 1.36 326 -2.17 3.171
##
## postpartum_sample = 4:
## contrast estimate SE df lower.CL upper.CL
## gram.neg.IMI.cow1 - gram.neg.IMI.cow0 -0.148 1.39 326 -2.87 2.578
##
## postpartum_sample = 5:
## contrast estimate SE df lower.CL upper.CL
## gram.neg.IMI.cow1 - gram.neg.IMI.cow0 0.615 1.42 326 -2.18 3.407
##
## postpartum_sample = 6:
## contrast estimate SE df lower.CL upper.CL
## gram.neg.IMI.cow1 - gram.neg.IMI.cow0 1.635 1.48 326 -1.29 4.556
##
## Results are averaged over the levels of: calving.nas, calving.sau.sslo
## Degrees-of-freedom method: containment
## Confidence level used: 0.95adj_means_12 <- as.data.frame(emmeans_12$emmeans)
adj_means_12## postpartum_sample = 1:
## gram.neg.IMI.cow emmean SE df lower.CL upper.CL
## 0 23.34270 3.093711 3 13.49713 33.18827
## 1 21.65706 3.318725 3 11.09540 32.21873
##
## postpartum_sample = 2:
## gram.neg.IMI.cow emmean SE df lower.CL upper.CL
## 0 25.24450 3.094065 3 15.39780 35.09119
## 1 25.01503 3.318466 3 14.45419 35.57587
##
## postpartum_sample = 3:
## gram.neg.IMI.cow emmean SE df lower.CL upper.CL
## 0 24.31372 3.095328 3 14.46300 34.16443
## 1 24.81514 3.331239 3 14.21365 35.41663
##
## postpartum_sample = 4:
## gram.neg.IMI.cow emmean SE df lower.CL upper.CL
## 0 23.36349 3.100857 3 13.49518 33.23180
## 1 23.21506 3.339960 3 12.58582 33.84431
##
## postpartum_sample = 5:
## gram.neg.IMI.cow emmean SE df lower.CL upper.CL
## 0 22.69904 3.097818 3 12.84040 32.55768
## 1 23.31451 3.355560 3 12.63562 33.99340
##
## postpartum_sample = 6:
## gram.neg.IMI.cow emmean SE df lower.CL upper.CL
## 0 20.49715 3.098898 3 10.63508 30.35923
## 1 22.13224 3.382845 3 11.36651 32.89796
##
## Results are averaged over the levels of: calving.nas, calving.sau.sslo
## Degrees-of-freedom method: containment
## Confidence level used: 0.95contrasts_12 <- as.data.frame(emmeans_12$contrasts) %>% select (postpartum_sample,p.value)
adj_means_12 <- adj_means_12 %>% left_join(contrasts_12, by="postpartum_sample")
adj_means_12 <- adj_means_12 %>% group_by(postpartum_sample) %>% mutate(max_milk = max(upper.CL)) %>% ungroup()
adj_means_12 <- adj_means_12 %>% mutate(p.value = ifelse(max_milk == upper.CL, p.value, 1))
milk_12 <- ggplot(adj_means_12, aes(x=postpartum_sample, y=emmean, group=gram.neg.IMI.cow, color=gram.neg.IMI.cow)) +
geom_line() +
geom_point()+
geom_errorbar(aes(ymin=emmean-SE, ymax=emmean+SE), width=.2,
position=position_dodge(0.05)) +
ylim(0,32) +
labs(x="Monthly postpartum test",y="Milk production (Kg/day)",title="Gram-negative bacteria",color="IMI") +
theme_classic() +
scale_color_viridis(discrete = TRUE) +
scale_color_manual(labels = c("No", "Yes"),
values = c("#009999", "#990000")) +
geom_text(aes(label = ifelse(p.value < 0.05, "*", "")),
vjust = -3, size = 5, color = "black")## Scale for colour is already present.
## Adding another scale for colour, which will replace the existing scale.model_12_adj_overall <- lme(fixed =milk_kg~ gram.neg.IMI.cow+calving.sau.sslo+calving.nas,random=~1|Site/CowId,correlation= corAR1(form = ~ps_sample),data=df.per.cow_dhi,na.action = na.omit)
Confint(model_12_adj_overall)## Estimate 2.5 % 97.5 %
## (Intercept) 24.374813199 18.349578 30.4000485
## gram.neg.IMI.cow1 -0.006239189 -1.838801 1.8263226
## calving.sau.sslo1 -1.725735452 -2.906934 -0.5445365
## calving.nas1 -0.432707127 -1.630841 0.7654267# ---- Others ----
model_13 <- lme(fixed =milk_kg~ other.IMI.cow*postpartum_sample,random=~1|Site/CowId,correlation= corAR1(form = ~ps_sample),data=df.per.cow_dhi,na.action = na.omit)
emmeans(model_13,revpairwise~other.IMI.cow|postpartum_sample)## $emmeans
## postpartum_sample = 1:
## other.IMI.cow emmean SE df lower.CL upper.CL
## 0 23.8 3.14 3 13.8 33.8
## 1 23.3 3.18 3 13.2 33.4
##
## postpartum_sample = 2:
## other.IMI.cow emmean SE df lower.CL upper.CL
## 0 25.3 3.14 3 15.3 35.3
## 1 26.2 3.18 3 16.1 36.4
##
## postpartum_sample = 3:
## other.IMI.cow emmean SE df lower.CL upper.CL
## 0 24.3 3.14 3 14.3 34.3
## 1 25.6 3.18 3 15.5 35.8
##
## postpartum_sample = 4:
## other.IMI.cow emmean SE df lower.CL upper.CL
## 0 23.0 3.15 3 13.0 33.1
## 1 25.0 3.19 3 14.8 35.1
##
## postpartum_sample = 5:
## other.IMI.cow emmean SE df lower.CL upper.CL
## 0 22.5 3.15 3 12.5 32.5
## 1 24.3 3.19 3 14.2 34.5
##
## postpartum_sample = 6:
## other.IMI.cow emmean SE df lower.CL upper.CL
## 0 20.6 3.15 3 10.6 30.6
## 1 21.8 3.19 3 11.7 32.0
##
## Degrees-of-freedom method: containment
## Confidence level used: 0.95
##
## $contrasts
## postpartum_sample = 1:
## contrast estimate SE df t.ratio p.value
## other.IMI.cow1 - other.IMI.cow0 -0.500 0.851 328 -0.587 0.5574
##
## postpartum_sample = 2:
## contrast estimate SE df t.ratio p.value
## other.IMI.cow1 - other.IMI.cow0 0.943 0.853 328 1.106 0.2695
##
## postpartum_sample = 3:
## contrast estimate SE df t.ratio p.value
## other.IMI.cow1 - other.IMI.cow0 1.355 0.876 328 1.547 0.1228
##
## postpartum_sample = 4:
## contrast estimate SE df t.ratio p.value
## other.IMI.cow1 - other.IMI.cow0 1.941 0.915 328 2.121 0.0347
##
## postpartum_sample = 5:
## contrast estimate SE df t.ratio p.value
## other.IMI.cow1 - other.IMI.cow0 1.846 0.903 328 2.044 0.0418
##
## postpartum_sample = 6:
## contrast estimate SE df t.ratio p.value
## other.IMI.cow1 - other.IMI.cow0 1.234 0.917 328 1.346 0.1792
##
## Degrees-of-freedom method: containmentmodel_13 <- lme(fixed =milk_kg~ other.IMI.cow*postpartum_sample+calving.nas+calving.sau.sslo,random=~1|Site/CowId,correlation= corAR1(form = ~ps_sample),data=df.per.cow_dhi,na.action = na.omit)
emmeans_13 <- emmeans(model_13,revpairwise~other.IMI.cow|postpartum_sample)
emmeans_13## $emmeans
## postpartum_sample = 1:
## other.IMI.cow emmean SE df lower.CL upper.CL
## 0 23.5 3.05 3 13.8 33.2
## 1 22.9 3.09 3 13.1 32.8
##
## postpartum_sample = 2:
## other.IMI.cow emmean SE df lower.CL upper.CL
## 0 25.0 3.05 3 15.3 34.7
## 1 25.8 3.10 3 16.0 35.7
##
## postpartum_sample = 3:
## other.IMI.cow emmean SE df lower.CL upper.CL
## 0 24.0 3.05 3 14.3 33.7
## 1 25.3 3.10 3 15.4 35.1
##
## postpartum_sample = 4:
## other.IMI.cow emmean SE df lower.CL upper.CL
## 0 22.7 3.06 3 13.0 32.5
## 1 24.6 3.10 3 14.7 34.4
##
## postpartum_sample = 5:
## other.IMI.cow emmean SE df lower.CL upper.CL
## 0 22.2 3.06 3 12.5 31.9
## 1 23.9 3.10 3 14.1 33.8
##
## postpartum_sample = 6:
## other.IMI.cow emmean SE df lower.CL upper.CL
## 0 20.3 3.06 3 10.6 30.0
## 1 21.4 3.11 3 11.6 31.3
##
## Results are averaged over the levels of: calving.nas, calving.sau.sslo
## Degrees-of-freedom method: containment
## Confidence level used: 0.95
##
## $contrasts
## postpartum_sample = 1:
## contrast estimate SE df t.ratio p.value
## other.IMI.cow1 - other.IMI.cow0 -0.616 0.856 326 -0.720 0.4721
##
## postpartum_sample = 2:
## contrast estimate SE df t.ratio p.value
## other.IMI.cow1 - other.IMI.cow0 0.838 0.857 326 0.978 0.3287
##
## postpartum_sample = 3:
## contrast estimate SE df t.ratio p.value
## other.IMI.cow1 - other.IMI.cow0 1.271 0.881 326 1.442 0.1502
##
## postpartum_sample = 4:
## contrast estimate SE df t.ratio p.value
## other.IMI.cow1 - other.IMI.cow0 1.855 0.920 326 2.016 0.0446
##
## postpartum_sample = 5:
## contrast estimate SE df t.ratio p.value
## other.IMI.cow1 - other.IMI.cow0 1.752 0.909 326 1.928 0.0547
##
## postpartum_sample = 6:
## contrast estimate SE df t.ratio p.value
## other.IMI.cow1 - other.IMI.cow0 1.151 0.922 326 1.249 0.2126
##
## Results are averaged over the levels of: calving.nas, calving.sau.sslo
## Degrees-of-freedom method: containmentconfint(emmeans_13)## $emmeans
## postpartum_sample = 1:
## other.IMI.cow emmean SE df lower.CL upper.CL
## 0 23.5 3.05 3 13.8 33.2
## 1 22.9 3.09 3 13.1 32.8
##
## postpartum_sample = 2:
## other.IMI.cow emmean SE df lower.CL upper.CL
## 0 25.0 3.05 3 15.3 34.7
## 1 25.8 3.10 3 16.0 35.7
##
## postpartum_sample = 3:
## other.IMI.cow emmean SE df lower.CL upper.CL
## 0 24.0 3.05 3 14.3 33.7
## 1 25.3 3.10 3 15.4 35.1
##
## postpartum_sample = 4:
## other.IMI.cow emmean SE df lower.CL upper.CL
## 0 22.7 3.06 3 13.0 32.5
## 1 24.6 3.10 3 14.7 34.4
##
## postpartum_sample = 5:
## other.IMI.cow emmean SE df lower.CL upper.CL
## 0 22.2 3.06 3 12.5 31.9
## 1 23.9 3.10 3 14.1 33.8
##
## postpartum_sample = 6:
## other.IMI.cow emmean SE df lower.CL upper.CL
## 0 20.3 3.06 3 10.6 30.0
## 1 21.4 3.11 3 11.6 31.3
##
## Results are averaged over the levels of: calving.nas, calving.sau.sslo
## Degrees-of-freedom method: containment
## Confidence level used: 0.95
##
## $contrasts
## postpartum_sample = 1:
## contrast estimate SE df lower.CL upper.CL
## other.IMI.cow1 - other.IMI.cow0 -0.616 0.856 326 -2.2996 1.07
##
## postpartum_sample = 2:
## contrast estimate SE df lower.CL upper.CL
## other.IMI.cow1 - other.IMI.cow0 0.838 0.857 326 -0.8477 2.52
##
## postpartum_sample = 3:
## contrast estimate SE df lower.CL upper.CL
## other.IMI.cow1 - other.IMI.cow0 1.271 0.881 326 -0.4627 3.00
##
## postpartum_sample = 4:
## contrast estimate SE df lower.CL upper.CL
## other.IMI.cow1 - other.IMI.cow0 1.855 0.920 326 0.0449 3.67
##
## postpartum_sample = 5:
## contrast estimate SE df lower.CL upper.CL
## other.IMI.cow1 - other.IMI.cow0 1.752 0.909 326 -0.0354 3.54
##
## postpartum_sample = 6:
## contrast estimate SE df lower.CL upper.CL
## other.IMI.cow1 - other.IMI.cow0 1.151 0.922 326 -0.6621 2.96
##
## Results are averaged over the levels of: calving.nas, calving.sau.sslo
## Degrees-of-freedom method: containment
## Confidence level used: 0.95adj_means_13 <- as.data.frame(emmeans_13$emmeans)
adj_means_13## postpartum_sample = 1:
## other.IMI.cow emmean SE df lower.CL upper.CL
## 0 23.52108 3.051067 3 13.81122 33.23094
## 1 22.90496 3.093956 3 13.05861 32.75130
##
## postpartum_sample = 2:
## other.IMI.cow emmean SE df lower.CL upper.CL
## 0 24.99919 3.050909 3 15.28984 34.70855
## 1 25.83767 3.095156 3 15.98751 35.68784
##
## postpartum_sample = 3:
## other.IMI.cow emmean SE df lower.CL upper.CL
## 0 23.98122 3.052484 3 14.26685 33.69559
## 1 25.25208 3.099747 3 15.38730 35.11686
##
## postpartum_sample = 4:
## other.IMI.cow emmean SE df lower.CL upper.CL
## 0 22.70855 3.061907 3 12.96419 32.45291
## 1 24.56376 3.104284 3 14.68455 34.44298
##
## postpartum_sample = 5:
## other.IMI.cow emmean SE df lower.CL upper.CL
## 0 22.18873 3.056399 3 12.46190 31.91555
## 1 23.94064 3.104424 3 14.06098 33.82031
##
## postpartum_sample = 6:
## other.IMI.cow emmean SE df lower.CL upper.CL
## 0 20.29240 3.058088 3 10.56020 30.02461
## 1 21.44327 3.106973 3 11.55550 31.33105
##
## Results are averaged over the levels of: calving.nas, calving.sau.sslo
## Degrees-of-freedom method: containment
## Confidence level used: 0.95contrasts_13 <- as.data.frame(emmeans_13$contrasts) %>% select (postpartum_sample,p.value)
adj_means_13 <- adj_means_13 %>% left_join(contrasts_13, by="postpartum_sample")
adj_means_13 <- adj_means_13 %>% group_by(postpartum_sample) %>% mutate(max_milk = max(upper.CL)) %>% ungroup()
adj_means_13 <- adj_means_13 %>% mutate(p.value = ifelse(max_milk == upper.CL, p.value, 1))
milk_13 <- ggplot(adj_means_13, aes(x=postpartum_sample, y=emmean, group=other.IMI.cow, color=other.IMI.cow)) +
geom_line() +
geom_point()+
geom_errorbar(aes(ymin=emmean-SE, ymax=emmean+SE), width=.2,
position=position_dodge(0.05)) +
ylim(0,32) +
labs(x="Monthly postpartum test",y="Milk production (Kg/day)",title="Others",color="IMI") +
theme_classic() +
scale_color_viridis(discrete = TRUE) +
scale_color_manual(labels = c("No", "Yes"),
values = c("#009999", "#990000")) +
geom_text(aes(label = ifelse(p.value < 0.05, "*", "")),
vjust = -3, size = 5, color = "black")## Scale for colour is already present.
## Adding another scale for colour, which will replace the existing scale.model_13_adj_overall <- lme(fixed =milk_kg~ other.IMI.cow+calving.sau.sslo+calving.nas,random=~1|Site/CowId,correlation= corAR1(form = ~ps_sample),data=df.per.cow_dhi,na.action = na.omit)
Confint(model_13_adj_overall)## Estimate 2.5 % 97.5 %
## (Intercept) 24.0502905 18.1079560 29.9926251
## other.IMI.cow1 0.8350230 -0.4155544 2.0856004
## calving.sau.sslo1 -1.6776553 -2.8585014 -0.4968093
## calving.nas1 -0.2824471 -1.4992564 0.9343622# ---- Bacillus ----
model_14 <- lme(fixed =milk_kg~ bacillus.IMI.cow*postpartum_sample,random=~1|Site/CowId,correlation= corAR1(form = ~ps_sample),data=df.per.cow_dhi,na.action = na.omit)
emmeans(model_14,revpairwise~bacillus.IMI.cow|postpartum_sample)## $emmeans
## postpartum_sample = 1:
## bacillus.IMI.cow emmean SE df lower.CL upper.CL
## 0 23.4 3.13 3 13.4 33.3
## 1 24.4 3.21 3 14.2 34.6
##
## postpartum_sample = 2:
## bacillus.IMI.cow emmean SE df lower.CL upper.CL
## 0 25.2 3.13 3 15.3 35.2
## 1 27.0 3.21 3 16.8 37.2
##
## postpartum_sample = 3:
## bacillus.IMI.cow emmean SE df lower.CL upper.CL
## 0 24.3 3.13 3 14.4 34.3
## 1 26.3 3.22 3 16.1 36.5
##
## postpartum_sample = 4:
## bacillus.IMI.cow emmean SE df lower.CL upper.CL
## 0 23.2 3.14 3 13.3 33.2
## 1 25.5 3.22 3 15.3 35.8
##
## postpartum_sample = 5:
## bacillus.IMI.cow emmean SE df lower.CL upper.CL
## 0 22.7 3.13 3 12.7 32.6
## 1 25.0 3.22 3 14.7 35.2
##
## postpartum_sample = 6:
## bacillus.IMI.cow emmean SE df lower.CL upper.CL
## 0 20.5 3.14 3 10.5 30.5
## 1 23.0 3.22 3 12.7 33.3
##
## Degrees-of-freedom method: containment
## Confidence level used: 0.95
##
## $contrasts
## postpartum_sample = 1:
## contrast estimate SE df t.ratio p.value
## bacillus.IMI.cow1 - bacillus.IMI.cow0 1.05 0.930 328 1.133 0.2580
##
## postpartum_sample = 2:
## contrast estimate SE df t.ratio p.value
## bacillus.IMI.cow1 - bacillus.IMI.cow0 1.75 0.936 328 1.871 0.0622
##
## postpartum_sample = 3:
## contrast estimate SE df t.ratio p.value
## bacillus.IMI.cow1 - bacillus.IMI.cow0 1.96 0.975 328 2.005 0.0458
##
## postpartum_sample = 4:
## contrast estimate SE df t.ratio p.value
## bacillus.IMI.cow1 - bacillus.IMI.cow0 2.29 1.000 328 2.292 0.0226
##
## postpartum_sample = 5:
## contrast estimate SE df t.ratio p.value
## bacillus.IMI.cow1 - bacillus.IMI.cow0 2.28 1.002 328 2.278 0.0234
##
## postpartum_sample = 6:
## contrast estimate SE df t.ratio p.value
## bacillus.IMI.cow1 - bacillus.IMI.cow0 2.50 1.012 328 2.474 0.0139
##
## Degrees-of-freedom method: containmentmodel_14 <- lme(fixed =milk_kg~ bacillus.IMI.cow*postpartum_sample+calving.nas+calving.sau.sslo,random=~1|Site/CowId,correlation= corAR1(form = ~ps_sample),data=df.per.cow_dhi,na.action = na.omit)
emmeans_14 <- emmeans(model_14,revpairwise~bacillus.IMI.cow|postpartum_sample)
emmeans_14## $emmeans
## postpartum_sample = 1:
## bacillus.IMI.cow emmean SE df lower.CL upper.CL
## 0 23.1 3.04 3 13.4 32.8
## 1 23.9 3.13 3 14.0 33.9
##
## postpartum_sample = 2:
## bacillus.IMI.cow emmean SE df lower.CL upper.CL
## 0 25.0 3.05 3 15.3 34.7
## 1 26.5 3.13 3 16.6 36.5
##
## postpartum_sample = 3:
## bacillus.IMI.cow emmean SE df lower.CL upper.CL
## 0 24.1 3.05 3 14.4 33.8
## 1 25.8 3.14 3 15.8 35.8
##
## postpartum_sample = 4:
## bacillus.IMI.cow emmean SE df lower.CL upper.CL
## 0 23.0 3.05 3 13.3 32.7
## 1 25.1 3.15 3 15.1 35.1
##
## postpartum_sample = 5:
## bacillus.IMI.cow emmean SE df lower.CL upper.CL
## 0 22.4 3.05 3 12.7 32.1
## 1 24.5 3.15 3 14.5 34.5
##
## postpartum_sample = 6:
## bacillus.IMI.cow emmean SE df lower.CL upper.CL
## 0 20.2 3.05 3 10.5 29.9
## 1 22.5 3.15 3 12.5 32.6
##
## Results are averaged over the levels of: calving.nas, calving.sau.sslo
## Degrees-of-freedom method: containment
## Confidence level used: 0.95
##
## $contrasts
## postpartum_sample = 1:
## contrast estimate SE df t.ratio p.value
## bacillus.IMI.cow1 - bacillus.IMI.cow0 0.819 0.935 326 0.876 0.3818
##
## postpartum_sample = 2:
## contrast estimate SE df t.ratio p.value
## bacillus.IMI.cow1 - bacillus.IMI.cow0 1.536 0.940 326 1.634 0.1032
##
## postpartum_sample = 3:
## contrast estimate SE df t.ratio p.value
## bacillus.IMI.cow1 - bacillus.IMI.cow0 1.761 0.979 326 1.799 0.0729
##
## postpartum_sample = 4:
## contrast estimate SE df t.ratio p.value
## bacillus.IMI.cow1 - bacillus.IMI.cow0 2.101 1.004 326 2.093 0.0371
##
## postpartum_sample = 5:
## contrast estimate SE df t.ratio p.value
## bacillus.IMI.cow1 - bacillus.IMI.cow0 2.085 1.006 326 2.073 0.0389
##
## postpartum_sample = 6:
## contrast estimate SE df t.ratio p.value
## bacillus.IMI.cow1 - bacillus.IMI.cow0 2.319 1.016 326 2.284 0.0230
##
## Results are averaged over the levels of: calving.nas, calving.sau.sslo
## Degrees-of-freedom method: containmentconfint(emmeans_14)## $emmeans
## postpartum_sample = 1:
## bacillus.IMI.cow emmean SE df lower.CL upper.CL
## 0 23.1 3.04 3 13.4 32.8
## 1 23.9 3.13 3 14.0 33.9
##
## postpartum_sample = 2:
## bacillus.IMI.cow emmean SE df lower.CL upper.CL
## 0 25.0 3.05 3 15.3 34.7
## 1 26.5 3.13 3 16.6 36.5
##
## postpartum_sample = 3:
## bacillus.IMI.cow emmean SE df lower.CL upper.CL
## 0 24.1 3.05 3 14.4 33.8
## 1 25.8 3.14 3 15.8 35.8
##
## postpartum_sample = 4:
## bacillus.IMI.cow emmean SE df lower.CL upper.CL
## 0 23.0 3.05 3 13.3 32.7
## 1 25.1 3.15 3 15.1 35.1
##
## postpartum_sample = 5:
## bacillus.IMI.cow emmean SE df lower.CL upper.CL
## 0 22.4 3.05 3 12.7 32.1
## 1 24.5 3.15 3 14.5 34.5
##
## postpartum_sample = 6:
## bacillus.IMI.cow emmean SE df lower.CL upper.CL
## 0 20.2 3.05 3 10.5 29.9
## 1 22.5 3.15 3 12.5 32.6
##
## Results are averaged over the levels of: calving.nas, calving.sau.sslo
## Degrees-of-freedom method: containment
## Confidence level used: 0.95
##
## $contrasts
## postpartum_sample = 1:
## contrast estimate SE df lower.CL upper.CL
## bacillus.IMI.cow1 - bacillus.IMI.cow0 0.819 0.935 326 -1.020 2.66
##
## postpartum_sample = 2:
## contrast estimate SE df lower.CL upper.CL
## bacillus.IMI.cow1 - bacillus.IMI.cow0 1.536 0.940 326 -0.313 3.38
##
## postpartum_sample = 3:
## contrast estimate SE df lower.CL upper.CL
## bacillus.IMI.cow1 - bacillus.IMI.cow0 1.761 0.979 326 -0.165 3.69
##
## postpartum_sample = 4:
## contrast estimate SE df lower.CL upper.CL
## bacillus.IMI.cow1 - bacillus.IMI.cow0 2.101 1.004 326 0.127 4.08
##
## postpartum_sample = 5:
## contrast estimate SE df lower.CL upper.CL
## bacillus.IMI.cow1 - bacillus.IMI.cow0 2.085 1.006 326 0.107 4.06
##
## postpartum_sample = 6:
## contrast estimate SE df lower.CL upper.CL
## bacillus.IMI.cow1 - bacillus.IMI.cow0 2.319 1.016 326 0.321 4.32
##
## Results are averaged over the levels of: calving.nas, calving.sau.sslo
## Degrees-of-freedom method: containment
## Confidence level used: 0.95adj_means_14 <- as.data.frame(emmeans_14$emmeans)
adj_means_14## postpartum_sample = 1:
## bacillus.IMI.cow emmean SE df lower.CL upper.CL
## 0 23.12474 3.044840 3 13.43470 32.81478
## 1 23.94341 3.130090 3 13.98206 33.90475
##
## postpartum_sample = 2:
## bacillus.IMI.cow emmean SE df lower.CL upper.CL
## 0 24.98658 3.045068 3 15.29581 34.67735
## 1 26.52245 3.131491 3 16.55664 36.48825
##
## postpartum_sample = 3:
## bacillus.IMI.cow emmean SE df lower.CL upper.CL
## 0 24.08029 3.046221 3 14.38585 33.77472
## 1 25.84079 3.141734 3 15.84239 35.83919
##
## postpartum_sample = 4:
## bacillus.IMI.cow emmean SE df lower.CL upper.CL
## 0 22.97119 3.053532 3 13.25349 32.68889
## 1 25.07219 3.145081 3 15.06314 35.08124
##
## postpartum_sample = 5:
## bacillus.IMI.cow emmean SE df lower.CL upper.CL
## 0 22.39812 3.049381 3 12.69362 32.10261
## 1 24.48354 3.148197 3 14.46457 34.50251
##
## postpartum_sample = 6:
## bacillus.IMI.cow emmean SE df lower.CL upper.CL
## 0 20.21681 3.051139 3 10.50672 29.92689
## 1 22.53602 3.149579 3 12.51265 32.55939
##
## Results are averaged over the levels of: calving.nas, calving.sau.sslo
## Degrees-of-freedom method: containment
## Confidence level used: 0.95contrasts_14 <- as.data.frame(emmeans_14$contrasts) %>% select (postpartum_sample,p.value)
adj_means_14 <- adj_means_14 %>% left_join(contrasts_14, by="postpartum_sample")
adj_means_14 <- adj_means_14 %>% group_by(postpartum_sample) %>% mutate(max_milk = max(upper.CL)) %>% ungroup()
adj_means_14 <- adj_means_14 %>% mutate(p.value = ifelse(max_milk == upper.CL, p.value, 1))
bacillus_title <- expression(paste(italic("Bacillus "),"spp."))
milk_14 <- ggplot(adj_means_14, aes(x=postpartum_sample, y=emmean, group=bacillus.IMI.cow, color=bacillus.IMI.cow)) +
geom_line() +
geom_point()+
geom_errorbar(aes(ymin=emmean-SE, ymax=emmean+SE), width=.2,
position=position_dodge(0.05)) +
ylim(0,32) +
labs(x="Monthly postpartum test",y="Milk production (Kg/day)",title=bacillus_title,color="IMI") +
theme_classic() +
scale_color_viridis(discrete = TRUE) +
scale_color_manual(labels = c("No", "Yes"),
values = c("#009999", "#990000")) +
geom_text(aes(label = ifelse(p.value < 0.05, "*", "")),
vjust = -3, size = 5, color = "black")## Scale for colour is already present.
## Adding another scale for colour, which will replace the existing scale.model_14_adj_overall <- lme(fixed =milk_kg~ bacillus.IMI.cow+calving.sau.sslo+calving.nas,random=~1|Site/CowId,correlation= corAR1(form = ~ps_sample),data=df.per.cow_dhi,na.action = na.omit)
Confint(model_14_adj_overall)## Estimate 2.5 % 97.5 %
## (Intercept) 23.9307560 17.9908922 29.8706198
## bacillus.IMI.cow1 1.6407200 0.3045264 2.9769136
## calving.sau.sslo1 -1.5524690 -2.7334855 -0.3714526
## calving.nas1 -0.2366099 -1.4368709 0.9636510# ---- Coryne ----
model_15 <- lme(fixed =milk_kg~ coryne.IMI.cow*postpartum_sample,random=~1|Site/CowId,correlation= corAR1(form = ~ps_sample),data=df.per.cow_dhi,na.action = na.omit)
emmeans(model_15,revpairwise~coryne.IMI.cow|postpartum_sample)## $emmeans
## postpartum_sample = 1:
## coryne.IMI.cow emmean SE df lower.CL upper.CL
## 0 23.8 3.21 3 13.6 34.0
## 1 22.1 3.34 3 11.5 32.7
##
## postpartum_sample = 2:
## coryne.IMI.cow emmean SE df lower.CL upper.CL
## 0 25.6 3.21 3 15.4 35.9
## 1 25.3 3.34 3 14.7 36.0
##
## postpartum_sample = 3:
## coryne.IMI.cow emmean SE df lower.CL upper.CL
## 0 24.5 3.21 3 14.3 34.7
## 1 25.8 3.35 3 15.2 36.5
##
## postpartum_sample = 4:
## coryne.IMI.cow emmean SE df lower.CL upper.CL
## 0 23.5 3.22 3 13.3 33.8
## 1 24.4 3.35 3 13.7 35.0
##
## postpartum_sample = 5:
## coryne.IMI.cow emmean SE df lower.CL upper.CL
## 0 23.2 3.21 3 13.0 33.4
## 1 22.7 3.35 3 12.1 33.4
##
## postpartum_sample = 6:
## coryne.IMI.cow emmean SE df lower.CL upper.CL
## 0 21.0 3.21 3 10.7 31.2
## 1 21.2 3.37 3 10.5 31.9
##
## Degrees-of-freedom method: containment
## Confidence level used: 0.95
##
## $contrasts
## postpartum_sample = 1:
## contrast estimate SE df t.ratio p.value
## coryne.IMI.cow1 - coryne.IMI.cow0 -1.725 1.10 328 -1.574 0.1165
##
## postpartum_sample = 2:
## contrast estimate SE df t.ratio p.value
## coryne.IMI.cow1 - coryne.IMI.cow0 -0.315 1.11 328 -0.284 0.7768
##
## postpartum_sample = 3:
## contrast estimate SE df t.ratio p.value
## coryne.IMI.cow1 - coryne.IMI.cow0 1.320 1.12 328 1.175 0.2409
##
## postpartum_sample = 4:
## contrast estimate SE df t.ratio p.value
## coryne.IMI.cow1 - coryne.IMI.cow0 0.832 1.15 328 0.722 0.4706
##
## postpartum_sample = 5:
## contrast estimate SE df t.ratio p.value
## coryne.IMI.cow1 - coryne.IMI.cow0 -0.468 1.15 328 -0.407 0.6839
##
## postpartum_sample = 6:
## contrast estimate SE df t.ratio p.value
## coryne.IMI.cow1 - coryne.IMI.cow0 0.229 1.20 328 0.191 0.8483
##
## Degrees-of-freedom method: containmentmodel_15 <- lme(fixed =milk_kg~ coryne.IMI.cow*postpartum_sample+calving.nas+calving.sau.sslo,random=~1|Site/CowId,correlation= corAR1(form = ~ps_sample),data=df.per.cow_dhi,na.action = na.omit)
emmeans_15 <- emmeans(model_15,revpairwise~coryne.IMI.cow|postpartum_sample)
emmeans_15## $emmeans
## postpartum_sample = 1:
## coryne.IMI.cow emmean SE df lower.CL upper.CL
## 0 23.5 3.10 3 13.6 33.3
## 1 21.8 3.24 3 11.4 32.1
##
## postpartum_sample = 2:
## coryne.IMI.cow emmean SE df lower.CL upper.CL
## 0 25.3 3.10 3 15.4 35.1
## 1 25.0 3.24 3 14.7 35.3
##
## postpartum_sample = 3:
## coryne.IMI.cow emmean SE df lower.CL upper.CL
## 0 24.1 3.10 3 14.3 34.0
## 1 25.5 3.25 3 15.1 35.8
##
## postpartum_sample = 4:
## coryne.IMI.cow emmean SE df lower.CL upper.CL
## 0 23.2 3.11 3 13.3 33.1
## 1 24.0 3.25 3 13.7 34.4
##
## postpartum_sample = 5:
## coryne.IMI.cow emmean SE df lower.CL upper.CL
## 0 22.8 3.11 3 12.9 32.7
## 1 22.4 3.25 3 12.0 32.7
##
## postpartum_sample = 6:
## coryne.IMI.cow emmean SE df lower.CL upper.CL
## 0 20.6 3.11 3 10.7 30.5
## 1 20.8 3.27 3 10.4 31.2
##
## Results are averaged over the levels of: calving.nas, calving.sau.sslo
## Degrees-of-freedom method: containment
## Confidence level used: 0.95
##
## $contrasts
## postpartum_sample = 1:
## contrast estimate SE df t.ratio p.value
## coryne.IMI.cow1 - coryne.IMI.cow0 -1.707 1.10 326 -1.556 0.1208
##
## postpartum_sample = 2:
## contrast estimate SE df t.ratio p.value
## coryne.IMI.cow1 - coryne.IMI.cow0 -0.277 1.11 326 -0.249 0.8033
##
## postpartum_sample = 3:
## contrast estimate SE df t.ratio p.value
## coryne.IMI.cow1 - coryne.IMI.cow0 1.342 1.13 326 1.192 0.2343
##
## postpartum_sample = 4:
## contrast estimate SE df t.ratio p.value
## coryne.IMI.cow1 - coryne.IMI.cow0 0.847 1.15 326 0.734 0.4636
##
## postpartum_sample = 5:
## contrast estimate SE df t.ratio p.value
## coryne.IMI.cow1 - coryne.IMI.cow0 -0.465 1.15 326 -0.403 0.6869
##
## postpartum_sample = 6:
## contrast estimate SE df t.ratio p.value
## coryne.IMI.cow1 - coryne.IMI.cow0 0.242 1.20 326 0.202 0.8403
##
## Results are averaged over the levels of: calving.nas, calving.sau.sslo
## Degrees-of-freedom method: containmentconfint(emmeans_15)## $emmeans
## postpartum_sample = 1:
## coryne.IMI.cow emmean SE df lower.CL upper.CL
## 0 23.5 3.10 3 13.6 33.3
## 1 21.8 3.24 3 11.4 32.1
##
## postpartum_sample = 2:
## coryne.IMI.cow emmean SE df lower.CL upper.CL
## 0 25.3 3.10 3 15.4 35.1
## 1 25.0 3.24 3 14.7 35.3
##
## postpartum_sample = 3:
## coryne.IMI.cow emmean SE df lower.CL upper.CL
## 0 24.1 3.10 3 14.3 34.0
## 1 25.5 3.25 3 15.1 35.8
##
## postpartum_sample = 4:
## coryne.IMI.cow emmean SE df lower.CL upper.CL
## 0 23.2 3.11 3 13.3 33.1
## 1 24.0 3.25 3 13.7 34.4
##
## postpartum_sample = 5:
## coryne.IMI.cow emmean SE df lower.CL upper.CL
## 0 22.8 3.11 3 12.9 32.7
## 1 22.4 3.25 3 12.0 32.7
##
## postpartum_sample = 6:
## coryne.IMI.cow emmean SE df lower.CL upper.CL
## 0 20.6 3.11 3 10.7 30.5
## 1 20.8 3.27 3 10.4 31.2
##
## Results are averaged over the levels of: calving.nas, calving.sau.sslo
## Degrees-of-freedom method: containment
## Confidence level used: 0.95
##
## $contrasts
## postpartum_sample = 1:
## contrast estimate SE df lower.CL upper.CL
## coryne.IMI.cow1 - coryne.IMI.cow0 -1.707 1.10 326 -3.866 0.452
##
## postpartum_sample = 2:
## contrast estimate SE df lower.CL upper.CL
## coryne.IMI.cow1 - coryne.IMI.cow0 -0.277 1.11 326 -2.462 1.909
##
## postpartum_sample = 3:
## contrast estimate SE df lower.CL upper.CL
## coryne.IMI.cow1 - coryne.IMI.cow0 1.342 1.13 326 -0.873 3.557
##
## postpartum_sample = 4:
## contrast estimate SE df lower.CL upper.CL
## coryne.IMI.cow1 - coryne.IMI.cow0 0.847 1.15 326 -1.425 3.119
##
## postpartum_sample = 5:
## contrast estimate SE df lower.CL upper.CL
## coryne.IMI.cow1 - coryne.IMI.cow0 -0.465 1.15 326 -2.730 1.801
##
## postpartum_sample = 6:
## contrast estimate SE df lower.CL upper.CL
## coryne.IMI.cow1 - coryne.IMI.cow0 0.242 1.20 326 -2.119 2.603
##
## Results are averaged over the levels of: calving.nas, calving.sau.sslo
## Degrees-of-freedom method: containment
## Confidence level used: 0.95adj_means_15 <- as.data.frame(emmeans_15$emmeans)
adj_means_15## postpartum_sample = 1:
## coryne.IMI.cow emmean SE df lower.CL upper.CL
## 0 23.46034 3.101375 3 13.59038 33.33030
## 1 21.75314 3.240671 3 11.43988 32.06640
##
## postpartum_sample = 2:
## coryne.IMI.cow emmean SE df lower.CL upper.CL
## 0 25.26237 3.101661 3 15.39150 35.13324
## 1 24.98545 3.244353 3 14.66047 35.31043
##
## postpartum_sample = 3:
## coryne.IMI.cow emmean SE df lower.CL upper.CL
## 0 24.12634 3.103244 3 14.25043 34.00224
## 1 25.46800 3.247048 3 15.13444 35.80155
##
## postpartum_sample = 4:
## coryne.IMI.cow emmean SE df lower.CL upper.CL
## 0 23.16975 3.109571 3 13.27371 33.06579
## 1 24.01721 3.252772 3 13.66544 34.36898
##
## postpartum_sample = 5:
## coryne.IMI.cow emmean SE df lower.CL upper.CL
## 0 22.82729 3.106291 3 12.94168 32.71289
## 1 22.36266 3.253276 3 12.00929 32.71604
##
## postpartum_sample = 6:
## coryne.IMI.cow emmean SE df lower.CL upper.CL
## 0 20.58823 3.107245 3 10.69960 30.47687
## 1 20.83034 3.269769 3 10.42447 31.23620
##
## Results are averaged over the levels of: calving.nas, calving.sau.sslo
## Degrees-of-freedom method: containment
## Confidence level used: 0.95contrasts_15 <- as.data.frame(emmeans_15$contrasts) %>% select (postpartum_sample,p.value)
adj_means_15 <- adj_means_15 %>% left_join(contrasts_15, by="postpartum_sample")
adj_means_15 <- adj_means_15 %>% group_by(postpartum_sample) %>% mutate(max_milk = max(upper.CL)) %>% ungroup()
adj_means_15 <- adj_means_15 %>% mutate(p.value = ifelse(max_milk == upper.CL, p.value, 1))
bacillus_title <- expression(paste(italic("Bacillus "),"spp."))
coryne_title <- expression(paste(italic("Corynebacterium "),"spp."))
milk_15 <- ggplot(adj_means_15, aes(x=postpartum_sample, y=emmean, group=coryne.IMI.cow, color=coryne.IMI.cow)) +
geom_line() +
geom_point()+
geom_errorbar(aes(ymin=emmean-SE, ymax=emmean+SE), width=.2,
position=position_dodge(0.05)) +
ylim(0,32) +
labs(x="Monthly postpartum test",y="Milk production (Kg/day)",title=coryne_title,color="IMI") +
theme_classic() +
scale_color_viridis(discrete = TRUE) +
scale_color_manual(labels = c("No", "Yes"),
values = c("#009999", "#990000")) +
geom_text(aes(label = ifelse(p.value < 0.05, "*", "")),
vjust = -3, size = 5, color = "black")## Scale for colour is already present.
## Adding another scale for colour, which will replace the existing scale.model_15_adj_overall <- lme(fixed =milk_kg~ coryne.IMI.cow+calving.sau.sslo+calving.nas,random=~1|Site/CowId,correlation= corAR1(form = ~ps_sample),data=df.per.cow_dhi,na.action = na.omit)
Confint(model_15_adj_overall)## Estimate 2.5 % 97.5 %
## (Intercept) 24.4096492 18.359195 30.4601035
## coryne.IMI.cow1 -0.2210869 -1.817931 1.3757573
## calving.sau.sslo1 -1.7228949 -2.904027 -0.5417626
## calving.nas1 -0.4562007 -1.666231 0.7538293library(ggpubr)
staph_figure <- ggarrange( milk_1 + rremove("ylab") + rremove("xlab"),
milk_2 + rremove("ylab") + rremove("xlab"),
milk_3 + rremove("ylab") + rremove("xlab"),
milk_4 + rremove("ylab") + rremove("xlab"),
milk_5 + rremove("ylab") + rremove("xlab"),
ncol = 2, nrow = 3, common.legend = TRUE,align = "h",labels = c("A", "B", "C","D","E","F","G"))
staph_figure_annotated <- annotate_figure(staph_figure,
bottom = text_grob("Monthly postpartum test", color = "black",
hjust = 0.75, x = 0.5),
left = text_grob("Milk production (Kg/day)", color = "black", rot = 90))
ggsave(plot = last_plot(),"./figures_article_2/staph_spp.png",width = 20, height = 30, units = "cm")
sslo_figure <- ggarrange(milk_6 + rremove("ylab") + rremove("xlab"),
milk_7 + rremove("ylab") + rremove("xlab"),
milk_8 + rremove("ylab") + rremove("xlab"),
milk_9 + rremove("ylab") + rremove("xlab"),
milk_10 + rremove("ylab") + rremove("xlab"),
milk_11 + rremove("ylab") + rremove("xlab"),
ncol = 2, nrow = 3, common.legend = TRUE,align = "h",labels = c("A", "B", "C","D","E","F","G","H","I","J","K","L"))
sslo_figure_annotated <- annotate_figure(sslo_figure,
bottom = text_grob("Monthly postpartum test", color = "black",
hjust = 0.75, x = 0.5),
left = text_grob("Milk production (Kg/day)", color = "black", rot = 90))
sslo_figure_annotated
ggsave(plot = last_plot(),"./figures_article_2/sslo.png",width = 20, height = 30, units = "cm")
gn_others_figure <- ggarrange(milk_12 + rremove("ylab") + rremove("xlab"),
milk_13 + rremove("ylab") + rremove("xlab"),
milk_14 + rremove("ylab") + rremove("xlab"),
milk_15 + rremove("ylab") + rremove("xlab"),
ncol = 2, nrow = 2, common.legend = TRUE,align = "h",labels = c("A", "B", "C","D","E","F","G","H","I","J","K","L"))
gn_others_figure_annotated <- annotate_figure(gn_others_figure,
bottom = text_grob("Postpartum test", color = "black",
hjust = 0.75, x = 0.5),
left = text_grob("Milk production (Kg/day)", color = "black", rot = 90))
gn_others_figure_annotated
ggsave(plot = last_plot(),"./figures_article_2/others_gn.png",width = 20, height = 20, units = "cm") 2.3 CA: Number of DHIA tests with high SCC
# ---- Staph aureus ----
#theme_set(theme(axis.text.x=element_blank(),theme_classic))
nhscc_1 <- glmer.nb(n_high_scc~as.factor(staph.aureus.IMI.cow.p)+(1|Site),data=df_dhi_cow)
summary(nhscc_1)## Generalized linear mixed model fit by maximum likelihood (Laplace
## Approximation) [glmerMod]
## Family: Negative Binomial(1.9256) ( log )
## Formula: n_high_scc ~ as.factor(staph.aureus.IMI.cow.p) + (1 | Site)
## Data: df_dhi_cow
##
## AIC BIC logLik deviance df.resid
## 1014.4 1033.5 -502.2 1004.4 328
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -1.0820 -0.6943 -0.4837 0.5208 6.4994
##
## Random effects:
## Groups Name Variance Std.Dev.
## Site (Intercept) 0.2725 0.522
## Number of obs: 333, groups: Site, 4
##
## Fixed effects:
## Estimate Std. Error z value Pr(>|z|)
## (Intercept) -0.07368 0.27420 -0.269 0.788
## as.factor(staph.aureus.IMI.cow.p)1 0.52493 0.31970 1.642 0.101
## as.factor(staph.aureus.IMI.cow.p)2 0.99756 0.14466 6.896 5.35e-12 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## (Intr) a.(..IMI..)1
## a.(..IMI..)1 -0.070
## a.(..IMI..)2 -0.149 0.183emmeans(nhscc_1,revpairwise~staph.aureus.IMI.cow.p,type="response")## $emmeans
## staph.aureus.IMI.cow.p response SE df asymp.LCL asymp.UCL
## 0 0.929 0.255 Inf 0.543 1.59
## 1 1.570 0.638 Inf 0.708 3.48
## 2 2.519 0.731 Inf 1.426 4.45
##
## Confidence level used: 0.95
## Intervals are back-transformed from the log scale
##
## $contrasts
## contrast ratio SE df null z.ratio
## staph.aureus.IMI.cow.p1 / staph.aureus.IMI.cow.p0 1.69 0.540 Inf 1 1.642
## staph.aureus.IMI.cow.p2 / staph.aureus.IMI.cow.p0 2.71 0.392 Inf 1 6.896
## staph.aureus.IMI.cow.p2 / staph.aureus.IMI.cow.p1 1.60 0.523 Inf 1 1.451
## p.value
## 0.2280
## <.0001
## 0.3151
##
## P value adjustment: tukey method for comparing a family of 3 estimates
## Tests are performed on the log scalenhscc_1_adj <- glmer.nb(n_high_scc~as.factor(staph.aureus.IMI.cow.p)+calving.nas+calving.sslo+(1|Site),data=df_dhi_cow,
family=poisson(link="log"))
summary(nhscc_1_adj)## Generalized linear mixed model fit by maximum likelihood (Laplace
## Approximation) [glmerMod]
## Family: Negative Binomial(2.3133) ( log )
## Formula: n_high_scc ~ as.factor(staph.aureus.IMI.cow.p) + calving.nas +
## calving.sslo + (1 | Site)
## Data: df_dhi_cow
##
## AIC BIC logLik deviance df.resid
## 1000.1 1026.7 -493.0 986.1 326
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -1.1731 -0.7209 -0.5672 0.4965 6.1943
##
## Random effects:
## Groups Name Variance Std.Dev.
## Site (Intercept) 0.1612 0.4015
## Number of obs: 333, groups: Site, 4
##
## Fixed effects:
## Estimate Std. Error z value Pr(>|z|)
## (Intercept) -0.3397 0.2294 -1.481 0.1387
## as.factor(staph.aureus.IMI.cow.p)1 0.4196 0.3172 1.323 0.1860
## as.factor(staph.aureus.IMI.cow.p)2 1.1263 0.1456 7.737 1.02e-14 ***
## calving.nas1 0.2987 0.1466 2.037 0.0416 *
## calving.sslo1 0.6322 0.1507 4.194 2.74e-05 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## (Intr) a.(..IMI..)1 a.(..IMI..)2 clvng.n1
## a.(..IMI..)1 -0.099
## a.(..IMI..)2 -0.265 0.204
## calving.ns1 -0.277 0.181 0.268
## calvng.ssl1 -0.228 -0.134 0.182 0.226exp(Confint(nhscc_1_adj))## Estimate 2.5 % 97.5 %
## (Intercept) 0.7120179 0.4542028 1.116174
## as.factor(staph.aureus.IMI.cow.p)1 1.5213525 0.8169388 2.833154
## as.factor(staph.aureus.IMI.cow.p)2 3.0841480 2.3185792 4.102499
## calving.nas1 1.3480740 1.0113662 1.796880
## calving.sslo1 1.8816910 1.4004044 2.528385emmeans_p1 <- emmeans(nhscc_1_adj,revpairwise~staph.aureus.IMI.cow.p,type="response")
performance::icc(nhscc_1_adj)## Warning: mu of 1.2 is too close to zero, estimate of random effect variances may
## be unreliable.## # Intraclass Correlation Coefficient
##
## Adjusted ICC: 0.166
## Unadjusted ICC: 0.131emmeans_p1## $emmeans
## staph.aureus.IMI.cow.p response SE df asymp.LCL asymp.UCL
## 0 1.13 0.251 Inf 0.735 1.75
## 1 1.73 0.639 Inf 0.835 3.56
## 2 3.50 0.872 Inf 2.145 5.70
##
## Results are averaged over the levels of: calving.nas, calving.sslo
## Confidence level used: 0.95
## Intervals are back-transformed from the log scale
##
## $contrasts
## contrast ratio SE df null z.ratio
## staph.aureus.IMI.cow.p1 / staph.aureus.IMI.cow.p0 1.52 0.483 Inf 1 1.323
## staph.aureus.IMI.cow.p2 / staph.aureus.IMI.cow.p0 3.08 0.449 Inf 1 7.737
## staph.aureus.IMI.cow.p2 / staph.aureus.IMI.cow.p1 2.03 0.651 Inf 1 2.202
## p.value
## 0.3825
## <.0001
## 0.0708
##
## Results are averaged over the levels of: calving.nas, calving.sslo
## P value adjustment: tukey method for comparing a family of 3 estimates
## Tests are performed on the log scaleadj_means_p1 <- as.data.frame(emmeans_p1$emmeans)
adj_means_p1## staph.aureus.IMI.cow.p response SE df asymp.LCL asymp.UCL
## 0 1.134023 0.2508948 Inf 0.7350205 1.749623
## 1 1.725249 0.6388647 Inf 0.8349287 3.564956
## 2 3.497495 0.8720493 Inf 2.1454703 5.701535
##
## Results are averaged over the levels of: calving.nas, calving.sslo
## Confidence level used: 0.95
## Intervals are back-transformed from the log scalenhscc_plot_1 <- ggplot(adj_means_p1, aes(x=staph.aureus.IMI.cow.p, y=response, fill=staph.aureus.IMI.cow.p)) +
geom_bar(stat = "identity") +
geom_errorbar(aes(ymin=response-SE, ymax=response+SE), width=.2,
position=position_dodge(0.05)) +
coord_flip() +
ylim(0,6) +
labs(y="n tests with high SCC",title="Staphylococcus aureus", fill="IMI type") +
scale_x_discrete(limits = rev) +
theme_classic() +
theme(plot.title=element_text(face="italic"),axis.text.y=element_blank(),axis.title.y = element_blank()) +
scale_fill_manual(labels = c("No IMI", "Transient IMI", "Persistent IMI"),
values = c("#009999", "#FF9933","#990000"))
# Add IRR and 95% CI
contrasts_p1 <- as.data.frame(exp(Confint(nhscc_1_adj)))
contrasts_p1$comparison <- row.names(contrasts_p1)
contrasts_p1 <- contrasts_p1 %>% filter(grepl("IMI.cow.p",comparison))
contrasts_p1 <- contrasts_p1 %>% rename("lower"=c(2),"upper"=c(3))
# Create a new data frame with the extra observation
extra_observation <- data.frame(comparison = "0",
Estimate = 1,
upper = 1,
lower = 1)
# Combine the extra observation with the original data frame
contrasts_p1_with_extra <- rbind(contrasts_p1, extra_observation) %>% arrange(comparison)
contrasts_p1_with_extra$staph.aureus.IMI.cow <- as.factor(row_number(contrasts_p1_with_extra))
nhscc_plot_1_b <- ggplot(contrasts_p1_with_extra, aes(x=staph.aureus.IMI.cow,y=Estimate,colour=staph.aureus.IMI.cow)) +
geom_point() +
geom_errorbar(aes(ymin=lower, ymax=upper), width=.2,
position=position_dodge(0.05)) +
ylim(0,5) +
coord_flip() +
scale_x_discrete(limits = rev) +
geom_hline(yintercept=1, linetype="dashed", size=0.5) +
theme_classic() +
theme(axis.text.y=element_blank(),axis.title.y = element_blank()) +
scale_colour_manual(labels = c("No IMI", "Transient IMI", "Persistent IMI"),
values = c("#009999", "#FF9933","#990000"))## Warning: Using `size` aesthetic for lines was deprecated in ggplot2 3.4.0.
## ℹ Please use `linewidth` instead.
## This warning is displayed once every 8 hours.
## Call `lifecycle::last_lifecycle_warnings()` to see where this warning was
## generated.# ---- nas ----
nhscc_2 <- glmer.nb(n_high_scc~as.factor(nas.IMI.cow.p)+(1|Site),data=df_dhi_cow)
summary(nhscc_2)## Generalized linear mixed model fit by maximum likelihood (Laplace
## Approximation) [glmerMod]
## Family: Negative Binomial(1.1879) ( log )
## Formula: n_high_scc ~ as.factor(nas.IMI.cow.p) + (1 | Site)
## Data: df_dhi_cow
##
## AIC BIC logLik deviance df.resid
## 1057.1 1076.1 -523.6 1047.1 328
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -0.9287 -0.6468 -0.3387 0.3807 5.3829
##
## Random effects:
## Groups Name Variance Std.Dev.
## Site (Intercept) 0.4266 0.6531
## Number of obs: 333, groups: Site, 4
##
## Fixed effects:
## Estimate Std. Error z value Pr(>|z|)
## (Intercept) 0.218621 0.342840 0.638 0.524
## as.factor(nas.IMI.cow.p)1 -0.004837 0.184691 -0.026 0.979
## as.factor(nas.IMI.cow.p)2 -0.029165 0.169170 -0.172 0.863
##
## Correlation of Fixed Effects:
## (Intr) a.(.IMI..)1
## a.(.IMI..)1 -0.158
## a.(.IMI..)2 -0.173 0.373emmeans(nhscc_2,revpairwise~nas.IMI.cow.p,type="response")## $emmeans
## nas.IMI.cow.p response SE df asymp.LCL asymp.UCL
## 0 1.24 0.427 Inf 0.636 2.44
## 1 1.24 0.449 Inf 0.608 2.52
## 2 1.21 0.429 Inf 0.603 2.42
##
## Confidence level used: 0.95
## Intervals are back-transformed from the log scale
##
## $contrasts
## contrast ratio SE df null z.ratio p.value
## nas.IMI.cow.p1 / nas.IMI.cow.p0 0.995 0.184 Inf 1 -0.026 0.9996
## nas.IMI.cow.p2 / nas.IMI.cow.p0 0.971 0.164 Inf 1 -0.172 0.9837
## nas.IMI.cow.p2 / nas.IMI.cow.p1 0.976 0.194 Inf 1 -0.123 0.9918
##
## P value adjustment: tukey method for comparing a family of 3 estimates
## Tests are performed on the log scalenhscc_2_adj <- glmer.nb(n_high_scc~as.factor(nas.IMI.cow.p)+calving.sau+calving.sslo+(1|Site),data=df_dhi_cow,
family=poisson(link="log"))
summary(nhscc_2_adj)## Generalized linear mixed model fit by maximum likelihood (Laplace
## Approximation) [glmerMod]
## Family: Negative Binomial(1.8672) ( log )
## Formula: n_high_scc ~ as.factor(nas.IMI.cow.p) + calving.sau + calving.sslo +
## (1 | Site)
## Data: df_dhi_cow
##
## AIC BIC logLik deviance df.resid
## 1014.9 1041.6 -500.5 1000.9 326
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -1.0816 -0.6823 -0.5750 0.4641 6.1270
##
## Random effects:
## Groups Name Variance Std.Dev.
## Site (Intercept) 0.1725 0.4153
## Number of obs: 333, groups: Site, 4
##
## Fixed effects:
## Estimate Std. Error z value Pr(>|z|)
## (Intercept) -0.2927 0.2422 -1.209 0.226837
## as.factor(nas.IMI.cow.p)1 0.2259 0.1719 1.314 0.188834
## as.factor(nas.IMI.cow.p)2 0.2729 0.1613 1.692 0.090614 .
## calving.sau1 0.9870 0.1544 6.391 1.65e-10 ***
## calving.sslo1 0.5907 0.1553 3.804 0.000142 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## (Intr) a.(.IMI..)1 a.(.IMI..)2 calvng.s1
## a.(.IMI..)1 -0.279
## a.(.IMI..)2 -0.315 0.420
## calving.sa1 -0.242 0.142 0.227
## calvng.ssl1 -0.247 0.192 0.221 0.155exp(Confint(nhscc_2_adj))## Estimate 2.5 % 97.5 %
## (Intercept) 0.7462164 0.4641763 1.199628
## as.factor(nas.IMI.cow.p)1 1.2534936 0.8948844 1.755809
## as.factor(nas.IMI.cow.p)2 1.3137751 0.9577304 1.802182
## calving.sau1 2.6831372 1.9823986 3.631573
## calving.sslo1 1.8052135 1.3315695 2.447334emmeans_p2 <- emmeans(nhscc_2_adj,revpairwise~nas.IMI.cow.p,type="response")
performance::icc(nhscc_2_adj)## Warning: mu of 1.2 is too close to zero, estimate of random effect variances may
## be unreliable.## # Intraclass Correlation Coefficient
##
## Adjusted ICC: 0.168
## Unadjusted ICC: 0.139emmeans_p2## $emmeans
## nas.IMI.cow.p response SE df asymp.LCL asymp.UCL
## 0 1.64 0.382 Inf 1.04 2.59
## 1 2.06 0.542 Inf 1.23 3.45
## 2 2.16 0.557 Inf 1.30 3.58
##
## Results are averaged over the levels of: calving.sau, calving.sslo
## Confidence level used: 0.95
## Intervals are back-transformed from the log scale
##
## $contrasts
## contrast ratio SE df null z.ratio p.value
## nas.IMI.cow.p1 / nas.IMI.cow.p0 1.25 0.216 Inf 1 1.314 0.3873
## nas.IMI.cow.p2 / nas.IMI.cow.p0 1.31 0.212 Inf 1 1.692 0.2082
## nas.IMI.cow.p2 / nas.IMI.cow.p1 1.05 0.188 Inf 1 0.261 0.9630
##
## Results are averaged over the levels of: calving.sau, calving.sslo
## P value adjustment: tukey method for comparing a family of 3 estimates
## Tests are performed on the log scaleadj_means_p2 <- as.data.frame(emmeans_p2$emmeans)
adj_means_p2## nas.IMI.cow.p response SE df asymp.LCL asymp.UCL
## 0 1.642292 0.3822744 Inf 1.040680 2.591693
## 1 2.058603 0.5423335 Inf 1.228360 3.450003
## 2 2.157603 0.5574379 Inf 1.300335 3.580039
##
## Results are averaged over the levels of: calving.sau, calving.sslo
## Confidence level used: 0.95
## Intervals are back-transformed from the log scalenhscc_plot_2 <- ggplot(adj_means_p2, aes(x=nas.IMI.cow.p, y=response, fill=nas.IMI.cow.p)) +
geom_bar(stat = "identity") +
geom_errorbar(aes(ymin=response-SE, ymax=response+SE), width=.2,
position=position_dodge(0.05)) +
coord_flip() +
ylim(0,6) +
labs(y="n tests with high SCC",title="NASM", fill="IMI type") +
scale_x_discrete(limits = rev) +
theme_classic() +
theme(axis.text.y=element_blank(),axis.title.y = element_blank()) +
scale_fill_manual(labels = c("No IMI", "Transient IMI", "Persistent IMI"),
values = c("#009999", "#FF9933","#990000"))
# Add IRR and 95% CI
contrasts_p2 <- as.data.frame(exp(Confint(nhscc_2_adj)))
contrasts_p2$comparison <- row.names(contrasts_p2)
contrasts_p2 <- contrasts_p2 %>% filter(grepl("IMI.cow.p",comparison))
contrasts_p2 <- contrasts_p2 %>% rename("lower"=c(2),"upper"=c(3))
# Create a new data frame with the extra observation
extra_observation <- data.frame(comparison = "0",
Estimate = 1,
upper = 1,
lower = 1)
# Combine the extra observation with the original data frame
contrasts_p2_with_extra <- rbind(contrasts_p2, extra_observation) %>% arrange(comparison)
contrasts_p2_with_extra$nas.IMI.cow <- as.factor(row_number(contrasts_p2_with_extra))
nhscc_plot_2_b <- ggplot(contrasts_p2_with_extra, aes(x=nas.IMI.cow,y=Estimate,colour=nas.IMI.cow)) +
geom_point() +
geom_errorbar(aes(ymin=lower, ymax=upper), width=.2,
position=position_dodge(0.05)) +
ylim(0,5) +
coord_flip() +
scale_x_discrete(limits = rev) +
geom_hline(yintercept=1, linetype="dashed", size=0.5) +
theme_classic() +
theme(axis.text.y=element_blank(),axis.title.y = element_blank()) +
scale_colour_manual(labels = c("No IMI", "Transient IMI", "Persistent IMI"),
values = c("#009999", "#FF9933","#990000"))
# ---- staph chromogenes ----
nhscc_3 <- glmer.nb(n_high_scc~ sch.IMI.cow.p+(1|Site),data=df_dhi_cow)
summary(nhscc_3)## Generalized linear mixed model fit by maximum likelihood (Laplace
## Approximation) [glmerMod]
## Family: Negative Binomial(1.1915) ( log )
## Formula: n_high_scc ~ sch.IMI.cow.p + (1 | Site)
## Data: df_dhi_cow
##
## AIC BIC logLik deviance df.resid
## 1056.9 1075.9 -523.4 1046.9 328
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -0.9365 -0.6580 -0.3422 0.4194 5.1473
##
## Random effects:
## Groups Name Variance Std.Dev.
## Site (Intercept) 0.4264 0.653
## Number of obs: 333, groups: Site, 4
##
## Fixed effects:
## Estimate Std. Error z value Pr(>|z|)
## (Intercept) 0.21767 0.33970 0.641 0.522
## sch.IMI.cow.p1 0.05770 0.22044 0.262 0.794
## sch.IMI.cow.p2 -0.06562 0.16914 -0.388 0.698
##
## Correlation of Fixed Effects:
## (Intr) s.IMI..1
## sch.IMI.c.1 -0.103
## sch.IMI.c.2 -0.130 0.219emmeans(nhscc_3,revpairwise~sch.IMI.cow.p,type="response")## $emmeans
## sch.IMI.cow.p response SE df asymp.LCL asymp.UCL
## 0 1.24 0.422 Inf 0.639 2.42
## 1 1.32 0.508 Inf 0.619 2.80
## 2 1.16 0.418 Inf 0.576 2.35
##
## Confidence level used: 0.95
## Intervals are back-transformed from the log scale
##
## $contrasts
## contrast ratio SE df null z.ratio p.value
## sch.IMI.cow.p1 / sch.IMI.cow.p0 1.059 0.234 Inf 1 0.262 0.9629
## sch.IMI.cow.p2 / sch.IMI.cow.p0 0.936 0.158 Inf 1 -0.388 0.9204
## sch.IMI.cow.p2 / sch.IMI.cow.p1 0.884 0.218 Inf 1 -0.500 0.8715
##
## P value adjustment: tukey method for comparing a family of 3 estimates
## Tests are performed on the log scalenhscc_3_adj <- glmer.nb(n_high_scc~ sch.IMI.cow.p+calving.sau+calving.sslo+(1|Site),data=df_dhi_cow)
summary(nhscc_3_adj)## Generalized linear mixed model fit by maximum likelihood (Laplace
## Approximation) [glmerMod]
## Family: Negative Binomial(1.8634) ( log )
## Formula: n_high_scc ~ sch.IMI.cow.p + calving.sau + calving.sslo + (1 |
## Site)
## Data: df_dhi_cow
##
## AIC BIC logLik deviance df.resid
## 1016.7 1043.3 -501.3 1002.7 326
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -1.0807 -0.6798 -0.5298 0.4310 5.9787
##
## Random effects:
## Groups Name Variance Std.Dev.
## Site (Intercept) 0.1935 0.4399
## Number of obs: 333, groups: Site, 4
##
## Fixed effects:
## Estimate Std. Error z value Pr(>|z|)
## (Intercept) -0.2160 0.2459 -0.878 0.379858
## sch.IMI.cow.p1 0.2034 0.2009 1.013 0.311232
## sch.IMI.cow.p2 0.1537 0.1578 0.974 0.330002
## calving.sau1 0.9509 0.1523 6.245 4.23e-10 ***
## calving.sslo1 0.5606 0.1538 3.646 0.000266 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## (Intr) s.IMI..1 s.IMI..2 calvng.s1
## sch.IMI.c.1 -0.170
## sch.IMI.c.2 -0.225 0.250
## calving.sa1 -0.201 0.063 0.161
## calvng.ssl1 -0.210 0.137 0.178 0.133exp(Confint(nhscc_3_adj))## Estimate 2.5 % 97.5 %
## (Intercept) 0.8057629 0.4975894 1.304799
## sch.IMI.cow.p1 1.2256049 0.8267006 1.816991
## sch.IMI.cow.p2 1.1661357 0.8559456 1.588737
## calving.sau1 2.5880955 1.9203365 3.488054
## calving.sslo1 1.7518036 1.2960075 2.367900emmeans_p3 <- emmeans(nhscc_3_adj,revpairwise~sch.IMI.cow.p,type="response")
performance::icc(nhscc_3_adj)## Warning: mu of 1.2 is too close to zero, estimate of random effect variances may
## be unreliable.## # Intraclass Correlation Coefficient
##
## Adjusted ICC: 0.185
## Unadjusted ICC: 0.156emmeans_p3## $emmeans
## sch.IMI.cow.p response SE df asymp.LCL asymp.UCL
## 0 1.72 0.414 Inf 1.07 2.75
## 1 2.10 0.624 Inf 1.18 3.76
## 2 2.00 0.544 Inf 1.17 3.41
##
## Results are averaged over the levels of: calving.sau, calving.sslo
## Confidence level used: 0.95
## Intervals are back-transformed from the log scale
##
## $contrasts
## contrast ratio SE df null z.ratio p.value
## sch.IMI.cow.p1 / sch.IMI.cow.p0 1.226 0.246 Inf 1 1.013 0.5688
## sch.IMI.cow.p2 / sch.IMI.cow.p0 1.166 0.184 Inf 1 0.974 0.5933
## sch.IMI.cow.p2 / sch.IMI.cow.p1 0.951 0.212 Inf 1 -0.224 0.9728
##
## Results are averaged over the levels of: calving.sau, calving.sslo
## P value adjustment: tukey method for comparing a family of 3 estimates
## Tests are performed on the log scaleadj_means_p3 <- as.data.frame(emmeans_p3$emmeans)
adj_means_p3## sch.IMI.cow.p response SE df asymp.LCL asymp.UCL
## 0 1.715695 0.4141869 Inf 1.068935 2.753779
## 1 2.102765 0.6239178 Inf 1.175515 3.761433
## 2 2.000733 0.5437399 Inf 1.174513 3.408164
##
## Results are averaged over the levels of: calving.sau, calving.sslo
## Confidence level used: 0.95
## Intervals are back-transformed from the log scalenhscc_plot_3 <- ggplot(adj_means_p3, aes(x=sch.IMI.cow.p, y=response, fill=sch.IMI.cow.p)) +
geom_bar(stat = "identity") +
geom_errorbar(aes(ymin=response-SE, ymax=response+SE), width=.2,
position=position_dodge(0.05)) +
coord_flip() +
ylim(0,6) +
labs(y="n tests with high SCC",title="Staphylococcus chromogenes", fill="IMI type") +
scale_x_discrete(limits = rev) +
theme_classic() +
theme(plot.title=element_text(face="italic"),axis.text.y=element_blank(),axis.title.y = element_blank()) +
scale_fill_manual(labels = c("No IMI", "Transient IMI", "Persistent IMI"),
values = c("#009999", "#FF9933","#990000"))
# Add IRR and 95% CI
contrasts_p3 <- as.data.frame(exp(Confint(nhscc_3_adj)))
contrasts_p3$comparison <- row.names(contrasts_p3)
contrasts_p3 <- contrasts_p3 %>% filter(grepl("IMI.cow.p",comparison))
contrasts_p3 <- contrasts_p3 %>% rename("lower"=c(2),"upper"=c(3))
# Create a new data frame with the extra observation
extra_observation <- data.frame(comparison = "0",
Estimate = 1,
upper = 1,
lower = 1)
# Combine the extra observation with the original data frame
contrasts_p3_with_extra <- rbind(contrasts_p3, extra_observation) %>% arrange(comparison)
contrasts_p3_with_extra$sch.IMI.cow <- as.factor(row_number(contrasts_p3_with_extra))
nhscc_plot_3_b <- ggplot(contrasts_p3_with_extra, aes(x=sch.IMI.cow,y=Estimate,colour=sch.IMI.cow)) +
geom_point() +
geom_errorbar(aes(ymin=lower, ymax=upper), width=.2,
position=position_dodge(0.05)) +
ylim(0,5) +
coord_flip() +
scale_x_discrete(limits = rev) +
geom_hline(yintercept=1, linetype="dashed", size=0.5) +
theme_classic() +
theme(axis.text.y=element_blank(),axis.title.y = element_blank()) +
scale_colour_manual(labels = c("No IMI", "Transient IMI", "Persistent IMI"),
values = c("#009999", "#FF9933","#990000"))
# ---- nas non chromogenes -----
nhscc_4 <- glmer.nb(n_high_scc~as.factor(nas.non.chrom.IMI.cow.p)+(1|Site),data=df_dhi_cow)
summary(nhscc_4)## Generalized linear mixed model fit by maximum likelihood (Laplace
## Approximation) [glmerMod]
## Family: Negative Binomial(1.1899) ( log )
## Formula: n_high_scc ~ as.factor(nas.non.chrom.IMI.cow.p) + (1 | Site)
## Data: df_dhi_cow
##
## AIC BIC logLik deviance df.resid
## 1056.7 1075.7 -523.3 1046.7 328
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -0.9446 -0.6655 -0.3770 0.3568 5.3736
##
## Random effects:
## Groups Name Variance Std.Dev.
## Site (Intercept) 0.4199 0.648
## Number of obs: 333, groups: Site, 4
##
## Fixed effects:
## Estimate Std. Error z value Pr(>|z|)
## (Intercept) 0.2104 0.3352 0.628 0.530
## as.factor(nas.non.chrom.IMI.cow.p)1 -0.0680 0.1960 -0.347 0.729
## as.factor(nas.non.chrom.IMI.cow.p)2 0.1382 0.2645 0.522 0.601
##
## Correlation of Fixed Effects:
## (Intr) a.(...IMI..)1
## a.(...IMI..)1 -0.090
## a.(...IMI..)2 -0.064 0.173emmeans(nhscc_4,revpairwise~nas.non.chrom.IMI.cow.p,type="response")## $emmeans
## nas.non.chrom.IMI.cow.p response SE df asymp.LCL asymp.UCL
## 0 1.23 0.414 Inf 0.640 2.38
## 1 1.15 0.430 Inf 0.555 2.39
## 2 1.42 0.586 Inf 0.630 3.19
##
## Confidence level used: 0.95
## Intervals are back-transformed from the log scale
##
## $contrasts
## contrast ratio SE df null
## nas.non.chrom.IMI.cow.p1 / nas.non.chrom.IMI.cow.p0 0.934 0.183 Inf 1
## nas.non.chrom.IMI.cow.p2 / nas.non.chrom.IMI.cow.p0 1.148 0.304 Inf 1
## nas.non.chrom.IMI.cow.p2 / nas.non.chrom.IMI.cow.p1 1.229 0.370 Inf 1
## z.ratio p.value
## -0.347 0.9358
## 0.522 0.8604
## 0.686 0.7718
##
## P value adjustment: tukey method for comparing a family of 3 estimates
## Tests are performed on the log scalenhscc_4_adj <- glmer.nb(n_high_scc~as.factor(nas.non.chrom.IMI.cow.p)+calving.sau+calving.sslo+(1|Site),data=df_dhi_cow)
summary(nhscc_4_adj)## Generalized linear mixed model fit by maximum likelihood (Laplace
## Approximation) [glmerMod]
## Family: Negative Binomial(1.9317) ( log )
## Formula: n_high_scc ~ as.factor(nas.non.chrom.IMI.cow.p) + calving.sau +
## calving.sslo + (1 | Site)
## Data: df_dhi_cow
##
## AIC BIC logLik deviance df.resid
## 1014.2 1040.9 -500.1 1000.2 326
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -1.0779 -0.6893 -0.6201 0.4502 6.7601
##
## Random effects:
## Groups Name Variance Std.Dev.
## Site (Intercept) 0.1706 0.413
## Number of obs: 333, groups: Site, 4
##
## Fixed effects:
## Estimate Std. Error z value Pr(>|z|)
## (Intercept) -0.2143 0.2298 -0.933 0.351061
## as.factor(nas.non.chrom.IMI.cow.p)1 0.1533 0.1773 0.865 0.387126
## as.factor(nas.non.chrom.IMI.cow.p)2 0.4690 0.2408 1.948 0.051416 .
## calving.sau1 0.9924 0.1533 6.473 9.6e-11 ***
## calving.sslo1 0.5460 0.1505 3.629 0.000285 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## (Intr) a.(...IMI..)1 a.(...IMI..)2 calvng.s1
## a.(...IMI..)1 -0.167
## a.(...IMI..)2 -0.136 0.205
## calving.sa1 -0.216 0.158 0.209
## calvng.ssl1 -0.193 0.095 0.073 0.130exp(Confint(nhscc_4_adj))## Estimate 2.5 % 97.5 %
## (Intercept) 0.8071326 0.5144762 1.266265
## as.factor(nas.non.chrom.IMI.cow.p)1 1.1656933 0.8235430 1.649994
## as.factor(nas.non.chrom.IMI.cow.p)2 1.5983862 0.9971218 2.562213
## calving.sau1 2.6976781 1.9975281 3.643236
## calving.sslo1 1.7263792 1.2854517 2.318551emmeans_p4 <- emmeans(nhscc_4_adj,revpairwise~nas.non.chrom.IMI.cow.p,type="response")
performance::icc(nhscc_4_adj)## Warning: mu of 1.2 is too close to zero, estimate of random effect variances may
## be unreliable.## # Intraclass Correlation Coefficient
##
## Adjusted ICC: 0.168
## Unadjusted ICC: 0.139emmeans_p4## $emmeans
## nas.non.chrom.IMI.cow.p response SE df asymp.LCL asymp.UCL
## 0 1.74 0.396 Inf 1.12 2.72
## 1 2.03 0.560 Inf 1.18 3.49
## 2 2.78 0.902 Inf 1.48 5.25
##
## Results are averaged over the levels of: calving.sau, calving.sslo
## Confidence level used: 0.95
## Intervals are back-transformed from the log scale
##
## $contrasts
## contrast ratio SE df null
## nas.non.chrom.IMI.cow.p1 / nas.non.chrom.IMI.cow.p0 1.17 0.207 Inf 1
## nas.non.chrom.IMI.cow.p2 / nas.non.chrom.IMI.cow.p0 1.60 0.385 Inf 1
## nas.non.chrom.IMI.cow.p2 / nas.non.chrom.IMI.cow.p1 1.37 0.368 Inf 1
## z.ratio p.value
## 0.865 0.6625
## 1.948 0.1254
## 1.177 0.4667
##
## Results are averaged over the levels of: calving.sau, calving.sslo
## P value adjustment: tukey method for comparing a family of 3 estimates
## Tests are performed on the log scaleadj_means_p4 <- as.data.frame(emmeans_p4$emmeans)
adj_means_p4## nas.non.chrom.IMI.cow.p response SE df asymp.LCL asymp.UCL
## 0 1.741839 0.3955068 Inf 1.116174 2.718218
## 1 2.030450 0.5604649 Inf 1.182049 3.487783
## 2 2.784132 0.9015661 Inf 1.475880 5.252046
##
## Results are averaged over the levels of: calving.sau, calving.sslo
## Confidence level used: 0.95
## Intervals are back-transformed from the log scaleadj_means_p4$nas.non.chrom.IMI.cow.p <- as.factor(adj_means_p4$nas.non.chrom.IMI.cow.p)
nhscc_plot_4 <- ggplot(adj_means_p4, aes(x=nas.non.chrom.IMI.cow.p, y=response, fill=nas.non.chrom.IMI.cow.p)) +
geom_bar(stat = "identity") +
geom_errorbar(aes(ymin=response-SE, ymax=response+SE), width=.2,
position=position_dodge(0.05)) +
coord_flip() +
ylim(0,6) +
labs(y="n tests with high SCC",title=nas_nch_title, fill="IMI type") +
scale_x_discrete(limits = rev) +
theme_classic() +
theme(axis.text.y=element_blank(),axis.title.y = element_blank()) +
scale_fill_manual(labels = c("No IMI", "Transient IMI", "Persistent IMI"),
values = c("#009999", "#FF9944","#990000"))
# Add IRR and 95% CI
contrasts_p4 <- as.data.frame(exp(Confint(nhscc_4_adj)))
contrasts_p4$comparison <- row.names(contrasts_p4)
contrasts_p4 <- contrasts_p4 %>% filter(grepl("IMI.cow.p",comparison))
contrasts_p4 <- contrasts_p4 %>% rename("lower"=c(2),"upper"=c(3))
# Create a new data frame with the extra observation
extra_observation <- data.frame(comparison = "0",
Estimate = 1,
upper = 1,
lower = 1)
# Combine the extra observation with the original data frame
contrasts_p4_with_extra <- rbind(contrasts_p4, extra_observation) %>% arrange(comparison)
contrasts_p4_with_extra$nas.non.chromogenes.IMI <- as.factor(row_number(contrasts_p4_with_extra))
nhscc_plot_4_b <- ggplot(contrasts_p4_with_extra, aes(x=nas.non.chromogenes.IMI,y=Estimate,colour=nas.non.chromogenes.IMI)) +
geom_point() +
geom_errorbar(aes(ymin=lower, ymax=upper), width=.2,
position=position_dodge(0.05)) +
ylim(0,5) +
coord_flip() +
scale_x_discrete(limits = rev) +
geom_hline(yintercept=1, linetype="dashed", size=0.5) +
theme_classic() +
theme(axis.text.y=element_blank(),axis.title.y = element_blank()) +
scale_colour_manual(labels = c("No IMI", "Transient IMI", "Persistent IMI"),
values = c("#009999", "#FF9933","#990000"))
# ---- Staph sp ----
nhscc_5 <- glmer.nb(n_high_scc~as.factor(staph.spp.IMI.cow.p)+(1|Site),data=df_dhi_cow)
summary(nhscc_5)## Generalized linear mixed model fit by maximum likelihood (Laplace
## Approximation) [glmerMod]
## Family: Negative Binomial(1.1931) ( log )
## Formula: n_high_scc ~ as.factor(staph.spp.IMI.cow.p) + (1 | Site)
## Data: df_dhi_cow
##
## AIC BIC logLik deviance df.resid
## 1056.5 1075.5 -523.2 1046.5 328
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -0.9451 -0.6567 -0.3788 0.3544 5.3635
##
## Random effects:
## Groups Name Variance Std.Dev.
## Site (Intercept) 0.4202 0.6482
## Number of obs: 333, groups: Site, 4
##
## Fixed effects:
## Estimate Std. Error z value Pr(>|z|)
## (Intercept) 0.2139 0.3349 0.639 0.523
## as.factor(staph.spp.IMI.cow.p)1 -0.1138 0.2120 -0.537 0.592
## as.factor(staph.spp.IMI.cow.p)2 0.1331 0.2632 0.506 0.613
##
## Correlation of Fixed Effects:
## (Intr) a.(..IMI..)1
## a.(..IMI..)1 -0.078
## a.(..IMI..)2 -0.060 0.149emmeans(nhscc_5,revpairwise~staph.spp.IMI.cow.p,type="response")## $emmeans
## staph.spp.IMI.cow.p response SE df asymp.LCL asymp.UCL
## 0 1.24 0.415 Inf 0.642 2.39
## 1 1.11 0.422 Inf 0.523 2.34
## 2 1.41 0.585 Inf 0.629 3.18
##
## Confidence level used: 0.95
## Intervals are back-transformed from the log scale
##
## $contrasts
## contrast ratio SE df null z.ratio
## staph.spp.IMI.cow.p1 / staph.spp.IMI.cow.p0 0.892 0.189 Inf 1 -0.537
## staph.spp.IMI.cow.p2 / staph.spp.IMI.cow.p0 1.142 0.301 Inf 1 0.506
## staph.spp.IMI.cow.p2 / staph.spp.IMI.cow.p1 1.280 0.400 Inf 1 0.790
## p.value
## 0.8533
## 0.8686
## 0.7091
##
## P value adjustment: tukey method for comparing a family of 3 estimates
## Tests are performed on the log scalenhscc_5_adj <- glmer.nb(n_high_scc~as.factor(staph.spp.IMI.cow.p)+calving.sau+calving.sslo+(1|Site),data=df_dhi_cow)
summary(nhscc_5_adj)## Generalized linear mixed model fit by maximum likelihood (Laplace
## Approximation) [glmerMod]
## Family: Negative Binomial(1.9167) ( log )
## Formula:
## n_high_scc ~ as.factor(staph.spp.IMI.cow.p) + calving.sau + calving.sslo +
## (1 | Site)
## Data: df_dhi_cow
##
## AIC BIC logLik deviance df.resid
## 1014.9 1041.6 -500.5 1000.9 326
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -1.0538 -0.6911 -0.6238 0.5508 6.6742
##
## Random effects:
## Groups Name Variance Std.Dev.
## Site (Intercept) 0.1825 0.4272
## Number of obs: 333, groups: Site, 4
##
## Fixed effects:
## Estimate Std. Error z value Pr(>|z|)
## (Intercept) -0.18625 0.23501 -0.793 0.428038
## as.factor(staph.spp.IMI.cow.p)1 0.02929 0.19132 0.153 0.878319
## as.factor(staph.spp.IMI.cow.p)2 0.43307 0.23947 1.808 0.070537 .
## calving.sau1 0.97479 0.15254 6.390 1.66e-10 ***
## calving.sslo1 0.53515 0.15025 3.562 0.000368 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## (Intr) a.(..IMI..)1 a.(..IMI..)2 calvng.s1
## a.(..IMI..)1 -0.128
## a.(..IMI..)2 -0.121 0.168
## calving.sa1 -0.200 0.106 0.196
## calvng.ssl1 -0.179 0.037 0.060 0.120exp(Confint(nhscc_5_adj))## Estimate 2.5 % 97.5 %
## (Intercept) 0.8300619 0.5236884 1.315673
## as.factor(staph.spp.IMI.cow.p)1 1.0297248 0.7077284 1.498220
## as.factor(staph.spp.IMI.cow.p)2 1.5419837 0.9643663 2.465571
## calving.sau1 2.6506048 1.9656300 3.574277
## calving.sslo1 1.7077068 1.2720978 2.292483emmeans_p5 <- emmeans(nhscc_5_adj,revpairwise~staph.spp.IMI.cow.p,type="response")
performance::icc(nhscc_5_adj)## Warning: mu of 1.2 is too close to zero, estimate of random effect variances may
## be unreliable.## # Intraclass Correlation Coefficient
##
## Adjusted ICC: 0.177
## Unadjusted ICC: 0.149emmeans_p5## $emmeans
## staph.spp.IMI.cow.p response SE df asymp.LCL asymp.UCL
## 0 1.77 0.413 Inf 1.12 2.79
## 1 1.82 0.526 Inf 1.03 3.21
## 2 2.72 0.893 Inf 1.43 5.18
##
## Results are averaged over the levels of: calving.sau, calving.sslo
## Confidence level used: 0.95
## Intervals are back-transformed from the log scale
##
## $contrasts
## contrast ratio SE df null z.ratio
## staph.spp.IMI.cow.p1 / staph.spp.IMI.cow.p0 1.03 0.197 Inf 1 0.153
## staph.spp.IMI.cow.p2 / staph.spp.IMI.cow.p0 1.54 0.369 Inf 1 1.808
## staph.spp.IMI.cow.p2 / staph.spp.IMI.cow.p1 1.50 0.420 Inf 1 1.441
## p.value
## 0.9872
## 0.1668
## 0.3200
##
## Results are averaged over the levels of: calving.sau, calving.sslo
## P value adjustment: tukey method for comparing a family of 3 estimates
## Tests are performed on the log scaleadj_means_p5 <- as.data.frame(emmeans_p5$emmeans)
adj_means_p5## staph.spp.IMI.cow.p response SE df asymp.LCL asymp.UCL
## 0 1.765996 0.4126282 Inf 1.117133 2.791737
## 1 1.818490 0.5264436 Inf 1.031080 3.207225
## 2 2.723137 0.8932602 Inf 1.431703 5.179476
##
## Results are averaged over the levels of: calving.sau, calving.sslo
## Confidence level used: 0.95
## Intervals are back-transformed from the log scalenhscc_plot_5 <- ggplot(adj_means_p5, aes(x=staph.spp.IMI.cow.p, y=response, fill=staph.spp.IMI.cow.p)) +
geom_bar(stat = "identity") +
geom_errorbar(aes(ymin=response-SE, ymax=response+SE), width=.2,
position=position_dodge(0.05)) +
coord_flip() +
ylim(0,6) +
labs(y="n tests with high SCC",title=staph_sp_title, fill="IMI type") +
scale_x_discrete(limits = rev) +
theme_classic() +
theme(axis.text.y=element_blank(),axis.title.y = element_blank()) +
scale_fill_manual(labels = c("No IMI", "Transient IMI", "Persistent IMI"),
values = c("#009999", "#FF9944","#990000"))
# Add IRR and 95% CI
contrasts_p5 <- as.data.frame(exp(Confint(nhscc_5_adj)))
contrasts_p5$comparison <- row.names(contrasts_p5)
contrasts_p5 <- contrasts_p5 %>% filter(grepl("IMI.cow.p",comparison))
contrasts_p5 <- contrasts_p5 %>% rename("lower"=c(2),"upper"=c(3))
# Create a new data frame with the extra observation
extra_observation <- data.frame(comparison = "0",
Estimate = 1,
upper = 1,
lower = 1)
# Combine the extra observation with the original data frame
contrasts_p5_with_extra <- rbind(contrasts_p5, extra_observation) %>% arrange(comparison)
contrasts_p5_with_extra$staph.sp.IMI.cow <- as.factor(row_number(contrasts_p5_with_extra))
nhscc_plot_5_b <- ggplot(contrasts_p5_with_extra, aes(x=staph.sp.IMI.cow,y=Estimate,colour=staph.sp.IMI.cow)) +
geom_point() +
geom_errorbar(aes(ymin=lower, ymax=upper), width=.2,
position=position_dodge(0.05)) +
ylim(0,5) +
coord_flip() +
scale_x_discrete(limits = rev) +
geom_hline(yintercept=1, linetype="dashed", size=0.5) +
theme_classic() +
theme(axis.text.y=element_blank(),axis.title.y = element_blank()) +
scale_colour_manual(labels = c("No IMI", "Transient IMI", "Persistent IMI"),
values = c("#009999", "#FF9933","#990000"))
# ---- sslo ----
nhscc_6 <- glmer.nb(n_high_scc~as.factor(strep.IMI.cow.p)+(1|Site),data=df_dhi_cow)
summary(nhscc_6)## Generalized linear mixed model fit by maximum likelihood (Laplace
## Approximation) [glmerMod]
## Family: Negative Binomial(1.3409) ( log )
## Formula: n_high_scc ~ as.factor(strep.IMI.cow.p) + (1 | Site)
## Data: df_dhi_cow
##
## AIC BIC logLik deviance df.resid
## 1043.9 1063.0 -517.0 1033.9 328
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -1.0262 -0.6844 -0.5306 0.4239 6.1268
##
## Random effects:
## Groups Name Variance Std.Dev.
## Site (Intercept) 0.3353 0.579
## Number of obs: 333, groups: Site, 4
##
## Fixed effects:
## Estimate Std. Error z value Pr(>|z|)
## (Intercept) 0.04808 0.30493 0.158 0.874710
## as.factor(strep.IMI.cow.p)1 0.24825 0.16976 1.462 0.143644
## as.factor(strep.IMI.cow.p)2 0.73632 0.20440 3.602 0.000315 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## (Intr) a.(.IMI..)1
## a.(.IMI..)1 -0.169
## a.(.IMI..)2 -0.125 0.283emmeans(nhscc_6,revpairwise~strep.IMI.cow.p,type="response")## $emmeans
## strep.IMI.cow.p response SE df asymp.LCL asymp.UCL
## 0 1.05 0.320 Inf 0.577 1.91
## 1 1.34 0.434 Inf 0.714 2.53
## 2 2.19 0.756 Inf 1.114 4.31
##
## Confidence level used: 0.95
## Intervals are back-transformed from the log scale
##
## $contrasts
## contrast ratio SE df null z.ratio p.value
## strep.IMI.cow.p1 / strep.IMI.cow.p0 1.28 0.218 Inf 1 1.462 0.3092
## strep.IMI.cow.p2 / strep.IMI.cow.p0 2.09 0.427 Inf 1 3.602 0.0009
## strep.IMI.cow.p2 / strep.IMI.cow.p1 1.63 0.368 Inf 1 2.162 0.0778
##
## P value adjustment: tukey method for comparing a family of 3 estimates
## Tests are performed on the log scalenhscc_6_adj <- glmer.nb(n_high_scc~as.factor(strep.IMI.cow.p)+calving.sau+calving.nas+(1|Site),data=df_dhi_cow)
summary(nhscc_6_adj)## Generalized linear mixed model fit by maximum likelihood (Laplace
## Approximation) [glmerMod]
## Family: Negative Binomial(2.1099) ( log )
## Formula: n_high_scc ~ as.factor(strep.IMI.cow.p) + calving.sau + calving.nas +
## (1 | Site)
## Data: df_dhi_cow
##
## AIC BIC logLik deviance df.resid
## 1007.1 1033.7 -496.5 993.1 326
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -1.1755 -0.7012 -0.5475 0.4410 5.8608
##
## Random effects:
## Groups Name Variance Std.Dev.
## Site (Intercept) 0.1494 0.3866
## Number of obs: 333, groups: Site, 4
##
## Fixed effects:
## Estimate Std. Error z value Pr(>|z|)
## (Intercept) -0.2706 0.2244 -1.206 0.2279
## as.factor(strep.IMI.cow.p)1 0.1768 0.1590 1.112 0.2662
## as.factor(strep.IMI.cow.p)2 0.8848 0.1859 4.759 1.94e-06 ***
## calving.sau1 1.0090 0.1517 6.652 2.89e-11 ***
## calving.nas1 0.2416 0.1463 1.652 0.0985 .
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## (Intr) a.(.IMI..)1 a.(.IMI..)2 clvng.s1
## a.(.IMI..)1 -0.236
## a.(.IMI..)2 -0.227 0.304
## calving.sa1 -0.220 -0.057 0.156
## calving.ns1 -0.277 0.162 0.179 0.235exp(Confint(nhscc_6_adj))## Estimate 2.5 % 97.5 %
## (Intercept) 0.7629539 0.4914839 1.184370
## as.factor(strep.IMI.cow.p)1 1.1934012 0.8738071 1.629886
## as.factor(strep.IMI.cow.p)2 2.4224277 1.6826927 3.487361
## calving.sau1 2.7429386 2.0375202 3.692583
## calving.nas1 1.2733438 0.9559540 1.696111emmeans_p6 <- emmeans(nhscc_6_adj,revpairwise~strep.IMI.cow.p,type="response")
performance::icc(nhscc_6_adj)## Warning: mu of 1.2 is too close to zero, estimate of random effect variances may
## be unreliable.## # Intraclass Correlation Coefficient
##
## Adjusted ICC: 0.153
## Unadjusted ICC: 0.123emmeans_p6## $emmeans
## strep.IMI.cow.p response SE df asymp.LCL asymp.UCL
## 0 1.43 0.311 Inf 0.93 2.19
## 1 1.70 0.411 Inf 1.06 2.73
## 2 3.45 0.929 Inf 2.04 5.85
##
## Results are averaged over the levels of: calving.sau, calving.nas
## Confidence level used: 0.95
## Intervals are back-transformed from the log scale
##
## $contrasts
## contrast ratio SE df null z.ratio p.value
## strep.IMI.cow.p1 / strep.IMI.cow.p0 1.19 0.190 Inf 1 1.112 0.5068
## strep.IMI.cow.p2 / strep.IMI.cow.p0 2.42 0.450 Inf 1 4.759 <.0001
## strep.IMI.cow.p2 / strep.IMI.cow.p1 2.03 0.415 Inf 1 3.460 0.0016
##
## Results are averaged over the levels of: calving.sau, calving.nas
## P value adjustment: tukey method for comparing a family of 3 estimates
## Tests are performed on the log scaleadj_means_p6 <- as.data.frame(emmeans_p6$emmeans)
adj_means_p6## strep.IMI.cow.p response SE df asymp.LCL asymp.UCL
## 0 1.425868 0.3107043 Inf 0.9302459 2.185549
## 1 1.701632 0.4109926 Inf 1.0599278 2.731839
## 2 3.454061 0.9293949 Inf 2.0384245 5.852824
##
## Results are averaged over the levels of: calving.sau, calving.nas
## Confidence level used: 0.95
## Intervals are back-transformed from the log scalenhscc_plot_6 <- ggplot(adj_means_p6, aes(x=strep.IMI.cow.p, y=response, fill=strep.IMI.cow.p)) +
geom_bar(stat = "identity") +
geom_errorbar(aes(ymin=response-SE, ymax=response+SE), width=.2,
position=position_dodge(0.06)) +
coord_flip() +
ylim(0,6) +
labs(y="n tests with high SCC",title="SSLO", fill="IMI type") +
scale_x_discrete(limits = rev) +
theme_classic() +
theme(axis.text.y=element_blank(),axis.title.y = element_blank()) +
scale_fill_manual(labels = c("No IMI", "Transient IMI", "Persistent IMI"),
values = c("#009999", "#FF9944","#990000"))
# Add IRR and 95% CI
contrasts_p6 <- as.data.frame(exp(Confint(nhscc_6_adj)))
contrasts_p6$comparison <- row.names(contrasts_p6)
contrasts_p6 <- contrasts_p6 %>% filter(grepl("IMI.cow.p",comparison))
contrasts_p6 <- contrasts_p6 %>% rename("lower"=c(2),"upper"=c(3))
# Create a new data frame with the extra observation
extra_observation <- data.frame(comparison = "0",
Estimate = 1,
upper = 1,
lower = 1)
# Combine the extra observation with the original data frame
contrasts_p6_with_extra <- rbind(contrasts_p6, extra_observation) %>% arrange(comparison)
contrasts_p6_with_extra$strep.IMI.cow <- as.factor(row_number(contrasts_p6_with_extra))
nhscc_plot_6_b <- ggplot(contrasts_p6_with_extra, aes(x=strep.IMI.cow,y=Estimate,colour=strep.IMI.cow)) +
geom_point() +
geom_errorbar(aes(ymin=lower, ymax=upper), width=.2,
position=position_dodge(0.05)) +
ylim(0,5) +
coord_flip() +
scale_x_discrete(limits = rev) +
geom_hline(yintercept=1, linetype="dashed", size=0.5) +
theme_classic() +
theme(axis.text.y=element_blank(),axis.title.y = element_blank()) +
scale_colour_manual(labels = c("No IMI", "Transient IMI", "Persistent IMI"),
values = c("#009999", "#FF9933","#990000"))
# ---- Strep genus ----
nhscc_7 <- glmer.nb(n_high_scc~as.factor(strep.genus.IMI.cow.p)+(1|Site),data=df_dhi_cow)
summary(nhscc_7)## Generalized linear mixed model fit by maximum likelihood (Laplace
## Approximation) [glmerMod]
## Family: Negative Binomial(1.3406) ( log )
## Formula: n_high_scc ~ as.factor(strep.genus.IMI.cow.p) + (1 | Site)
## Data: df_dhi_cow
##
## AIC BIC logLik deviance df.resid
## 1045.9 1064.9 -517.9 1035.9 328
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -1.0187 -0.6900 -0.5313 0.3946 5.6287
##
## Random effects:
## Groups Name Variance Std.Dev.
## Site (Intercept) 0.2977 0.5456
## Number of obs: 333, groups: Site, 4
##
## Fixed effects:
## Estimate Std. Error z value Pr(>|z|)
## (Intercept) 0.08445 0.28583 0.295 0.76764
## as.factor(strep.genus.IMI.cow.p)1 0.40528 0.20928 1.937 0.05280 .
## as.factor(strep.genus.IMI.cow.p)2 0.66899 0.21287 3.143 0.00167 **
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## (Intr) a.(..IMI..)1
## a.(..IMI..)1 -0.104
## a.(..IMI..)2 -0.101 0.234emmeans(nhscc_7,revpairwise~strep.genus.IMI.cow.p,type="response")## $emmeans
## strep.genus.IMI.cow.p response SE df asymp.LCL asymp.UCL
## 0 1.09 0.311 Inf 0.621 1.91
## 1 1.63 0.549 Inf 0.844 3.15
## 2 2.12 0.719 Inf 1.094 4.13
##
## Confidence level used: 0.95
## Intervals are back-transformed from the log scale
##
## $contrasts
## contrast ratio SE df null z.ratio
## strep.genus.IMI.cow.p1 / strep.genus.IMI.cow.p0 1.50 0.314 Inf 1 1.937
## strep.genus.IMI.cow.p2 / strep.genus.IMI.cow.p0 1.95 0.416 Inf 1 3.143
## strep.genus.IMI.cow.p2 / strep.genus.IMI.cow.p1 1.30 0.340 Inf 1 1.009
## p.value
## 0.1285
## 0.0048
## 0.5709
##
## P value adjustment: tukey method for comparing a family of 3 estimates
## Tests are performed on the log scalenhscc_7_adj <- glmer.nb(n_high_scc~as.factor(strep.genus.IMI.cow.p)+calving.sau+calving.nas+(1|Site),data=df_dhi_cow)
summary(nhscc_7_adj)## Generalized linear mixed model fit by maximum likelihood (Laplace
## Approximation) [glmerMod]
## Family: Negative Binomial(2.0238) ( log )
## Formula: n_high_scc ~ as.factor(strep.genus.IMI.cow.p) + calving.sau +
## calving.nas + (1 | Site)
## Data: df_dhi_cow
##
## AIC BIC logLik deviance df.resid
## 1012.7 1039.3 -499.3 998.7 326
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -1.1505 -0.7235 -0.4997 0.4101 5.4245
##
## Random effects:
## Groups Name Variance Std.Dev.
## Site (Intercept) 0.1384 0.372
## Number of obs: 333, groups: Site, 4
##
## Fixed effects:
## Estimate Std. Error z value Pr(>|z|)
## (Intercept) -0.2197 0.2136 -1.029 0.304
## as.factor(strep.genus.IMI.cow.p)1 0.2451 0.1970 1.244 0.213
## as.factor(strep.genus.IMI.cow.p)2 0.8019 0.1952 4.108 4.00e-05 ***
## calving.sau1 0.9724 0.1539 6.317 2.66e-10 ***
## calving.nas1 0.2400 0.1472 1.630 0.103
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## (Intr) a.(..IMI..)1 a.(..IMI..)2 clvng.s1
## a.(..IMI..)1 -0.138
## a.(..IMI..)2 -0.192 0.249
## calving.sa1 -0.223 -0.130 0.133
## calving.ns1 -0.275 0.146 0.186 0.227exp(Confint(nhscc_7_adj))## Estimate 2.5 % 97.5 %
## (Intercept) 0.8027626 0.5282110 1.220020
## as.factor(strep.genus.IMI.cow.p)1 1.2777709 0.8685330 1.879835
## as.factor(strep.genus.IMI.cow.p)2 2.2298341 1.5208605 3.269307
## calving.sau1 2.6441797 1.9555850 3.575240
## calving.nas1 1.2712131 0.9526318 1.696335emmeans_p7 <- emmeans(nhscc_7_adj,revpairwise~strep.genus.IMI.cow.p,type="response")
performance::icc(nhscc_7_adj)## Warning: mu of 1.2 is too close to zero, estimate of random effect variances may
## be unreliable.## # Intraclass Correlation Coefficient
##
## Adjusted ICC: 0.142
## Unadjusted ICC: 0.116emmeans_p7## $emmeans
## strep.genus.IMI.cow.p response SE df asymp.LCL asymp.UCL
## 0 1.47 0.307 Inf 0.978 2.22
## 1 1.88 0.501 Inf 1.115 3.17
## 2 3.28 0.899 Inf 1.919 5.61
##
## Results are averaged over the levels of: calving.sau, calving.nas
## Confidence level used: 0.95
## Intervals are back-transformed from the log scale
##
## $contrasts
## contrast ratio SE df null z.ratio
## strep.genus.IMI.cow.p1 / strep.genus.IMI.cow.p0 1.28 0.252 Inf 1 1.244
## strep.genus.IMI.cow.p2 / strep.genus.IMI.cow.p0 2.23 0.435 Inf 1 4.108
## strep.genus.IMI.cow.p2 / strep.genus.IMI.cow.p1 1.75 0.420 Inf 1 2.316
## p.value
## 0.4270
## 0.0001
## 0.0536
##
## Results are averaged over the levels of: calving.sau, calving.nas
## P value adjustment: tukey method for comparing a family of 3 estimates
## Tests are performed on the log scaleadj_means_p7 <- as.data.frame(emmeans_p7$emmeans)
adj_means_p7## strep.genus.IMI.cow.p response SE df asymp.LCL asymp.UCL
## 0 1.471776 0.3072060 Inf 0.9776182 2.215717
## 1 1.880593 0.5011662 Inf 1.1154624 3.170550
## 2 3.281817 0.8987324 Inf 1.9187174 5.613292
##
## Results are averaged over the levels of: calving.sau, calving.nas
## Confidence level used: 0.95
## Intervals are back-transformed from the log scaleadj_means_p7$strep.genus.IMI.cow.p <- as.factor(adj_means_p7$strep.genus.IMI.cow.p)
nhscc_plot_7 <- ggplot(adj_means_p7, aes(x=strep.genus.IMI.cow.p, y=response, fill=strep.genus.IMI.cow.p)) +
geom_bar(stat = "identity") +
geom_errorbar(aes(ymin=response-SE, ymax=response+SE), width=.2,
position=position_dodge(0.06)) +
coord_flip() +
ylim(0,6) +
labs(y="n tests with high SCC",title=strep_genus_title, fill="IMI type") +
scale_x_discrete(limits = rev) +
theme_classic() +
theme(axis.text.y=element_blank(),axis.title.y = element_blank()) +
scale_fill_manual(labels = c("No IMI", "Transient IMI", "Persistent IMI"),
values = c("#009999", "#FF9944","#990000"))
# Add IRR and 95% CI
contrasts_p7 <- as.data.frame(exp(Confint(nhscc_7_adj)))
contrasts_p7$comparison <- row.names(contrasts_p7)
contrasts_p7 <- contrasts_p7 %>% filter(grepl("IMI.cow.p",comparison))
contrasts_p7 <- contrasts_p7 %>% rename("lower"=c(2),"upper"=c(3))
# Create a new data frame with the extra observation
extra_observation <- data.frame(comparison = "0",
Estimate = 1,
upper = 1,
lower = 1)
# Combine the extra observation with the original data frame
contrasts_p7_with_extra <- rbind(contrasts_p7, extra_observation) %>% arrange(comparison)
contrasts_p7_with_extra$strep.genus.IMI.cow <- as.factor(row_number(contrasts_p7_with_extra))
nhscc_plot_7_b <- ggplot(contrasts_p7_with_extra, aes(x=strep.genus.IMI.cow,y=Estimate,colour=strep.genus.IMI.cow)) +
geom_point() +
geom_errorbar(aes(ymin=lower, ymax=upper), width=.2,
position=position_dodge(0.05)) +
ylim(0,5) +
coord_flip() +
scale_x_discrete(limits = rev) +
geom_hline(yintercept=1, linetype="dashed", size=0.5) +
theme_classic() +
theme(axis.text.y=element_blank(),axis.title.y = element_blank()) +
scale_colour_manual(labels = c("No IMI", "Transient IMI", "Persistent IMI"),
values = c("#009999", "#FF9933","#990000"))
# ---- Strep dys -----
nhscc_8 <- glmer.nb(n_high_scc~as.factor(strep.dys.IMI.cow.p)+calving.sau+calving.nas+(1|Site),data=df_dhi_cow)
summary(nhscc_8)## Generalized linear mixed model fit by maximum likelihood (Laplace
## Approximation) [glmerMod]
## Family: Negative Binomial(2.0014) ( log )
## Formula:
## n_high_scc ~ as.factor(strep.dys.IMI.cow.p) + calving.sau + calving.nas +
## (1 | Site)
## Data: df_dhi_cow
##
## AIC BIC logLik deviance df.resid
## 1014.5 1041.2 -500.3 1000.5 326
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -1.1090 -0.6891 -0.4627 0.4176 5.3431
##
## Random effects:
## Groups Name Variance Std.Dev.
## Site (Intercept) 0.1626 0.4032
## Number of obs: 333, groups: Site, 4
##
## Fixed effects:
## Estimate Std. Error z value Pr(>|z|)
## (Intercept) -0.1946 0.2262 -0.860 0.38952
## as.factor(strep.dys.IMI.cow.p)1 0.2566 0.2370 1.083 0.27894
## as.factor(strep.dys.IMI.cow.p)2 0.8397 0.2194 3.827 0.00013 ***
## calving.sau1 0.9944 0.1536 6.473 9.61e-11 ***
## calving.nas1 0.2516 0.1477 1.703 0.08855 .
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## (Intr) a.(..IMI..)1 a.(..IMI..)2 clvng.s1
## a.(..IMI..)1 -0.083
## a.(..IMI..)2 -0.160 0.148
## calving.sa1 -0.227 -0.055 0.174
## calving.ns1 -0.255 0.087 0.212 0.258emmeans(nhscc_8,revpairwise~strep.dys.IMI.cow.p,type="response")## $emmeans
## strep.dys.IMI.cow.p response SE df asymp.LCL asymp.UCL
## 0 1.53 0.340 Inf 0.994 2.37
## 1 1.98 0.619 Inf 1.076 3.66
## 2 3.55 1.089 Inf 1.949 6.48
##
## Results are averaged over the levels of: calving.sau, calving.nas
## Confidence level used: 0.95
## Intervals are back-transformed from the log scale
##
## $contrasts
## contrast ratio SE df null z.ratio
## strep.dys.IMI.cow.p1 / strep.dys.IMI.cow.p0 1.29 0.306 Inf 1 1.083
## strep.dys.IMI.cow.p2 / strep.dys.IMI.cow.p0 2.32 0.508 Inf 1 3.827
## strep.dys.IMI.cow.p2 / strep.dys.IMI.cow.p1 1.79 0.534 Inf 1 1.956
## p.value
## 0.5248
## 0.0004
## 0.1234
##
## Results are averaged over the levels of: calving.sau, calving.nas
## P value adjustment: tukey method for comparing a family of 3 estimates
## Tests are performed on the log scalenhscc_8_adj <- glmer.nb(n_high_scc~as.factor(strep.dys.IMI.cow.p)+calving.sau+calving.nas+(1|Site),data=df_dhi_cow)
summary(nhscc_8_adj)## Generalized linear mixed model fit by maximum likelihood (Laplace
## Approximation) [glmerMod]
## Family: Negative Binomial(2.0014) ( log )
## Formula:
## n_high_scc ~ as.factor(strep.dys.IMI.cow.p) + calving.sau + calving.nas +
## (1 | Site)
## Data: df_dhi_cow
##
## AIC BIC logLik deviance df.resid
## 1014.5 1041.2 -500.3 1000.5 326
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -1.1090 -0.6891 -0.4627 0.4176 5.3431
##
## Random effects:
## Groups Name Variance Std.Dev.
## Site (Intercept) 0.1626 0.4032
## Number of obs: 333, groups: Site, 4
##
## Fixed effects:
## Estimate Std. Error z value Pr(>|z|)
## (Intercept) -0.1946 0.2262 -0.860 0.38952
## as.factor(strep.dys.IMI.cow.p)1 0.2566 0.2370 1.083 0.27894
## as.factor(strep.dys.IMI.cow.p)2 0.8397 0.2194 3.827 0.00013 ***
## calving.sau1 0.9944 0.1536 6.473 9.61e-11 ***
## calving.nas1 0.2516 0.1477 1.703 0.08855 .
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## (Intr) a.(..IMI..)1 a.(..IMI..)2 clvng.s1
## a.(..IMI..)1 -0.083
## a.(..IMI..)2 -0.160 0.148
## calving.sa1 -0.227 -0.055 0.174
## calving.ns1 -0.255 0.087 0.212 0.258exp(Confint(nhscc_8_adj))## Estimate 2.5 % 97.5 %
## (Intercept) 0.8231469 0.5283906 1.282329
## as.factor(strep.dys.IMI.cow.p)1 1.2925700 0.8122575 2.056906
## as.factor(strep.dys.IMI.cow.p)2 2.3155924 1.5063371 3.559607
## calving.sau1 2.7030657 2.0002996 3.652735
## calving.nas1 1.2860300 0.9627723 1.717824emmeans_p8 <- emmeans(nhscc_8_adj,revpairwise~strep.dys.IMI.cow.p,type="response")
performance::icc(nhscc_8_adj)## Warning: mu of 1.2 is too close to zero, estimate of random effect variances may
## be unreliable.## # Intraclass Correlation Coefficient
##
## Adjusted ICC: 0.163
## Unadjusted ICC: 0.135emmeans_p8## $emmeans
## strep.dys.IMI.cow.p response SE df asymp.LCL asymp.UCL
## 0 1.53 0.340 Inf 0.994 2.37
## 1 1.98 0.619 Inf 1.076 3.66
## 2 3.55 1.089 Inf 1.949 6.48
##
## Results are averaged over the levels of: calving.sau, calving.nas
## Confidence level used: 0.95
## Intervals are back-transformed from the log scale
##
## $contrasts
## contrast ratio SE df null z.ratio
## strep.dys.IMI.cow.p1 / strep.dys.IMI.cow.p0 1.29 0.306 Inf 1 1.083
## strep.dys.IMI.cow.p2 / strep.dys.IMI.cow.p0 2.32 0.508 Inf 1 3.827
## strep.dys.IMI.cow.p2 / strep.dys.IMI.cow.p1 1.79 0.534 Inf 1 1.956
## p.value
## 0.5248
## 0.0004
## 0.1234
##
## Results are averaged over the levels of: calving.sau, calving.nas
## P value adjustment: tukey method for comparing a family of 3 estimates
## Tests are performed on the log scaleadj_means_p8 <- as.data.frame(emmeans_p8$emmeans)
adj_means_p8## strep.dys.IMI.cow.p response SE df asymp.LCL asymp.UCL
## 0 1.534727 0.3398207 Inf 0.9943913 2.368673
## 1 1.983742 0.6188508 Inf 1.0763247 3.656177
## 2 3.553803 1.0893428 Inf 1.9488370 6.480538
##
## Results are averaged over the levels of: calving.sau, calving.nas
## Confidence level used: 0.95
## Intervals are back-transformed from the log scalenhscc_plot_8 <- ggplot(adj_means_p8, aes(x=strep.dys.IMI.cow.p, y=response, fill=strep.dys.IMI.cow.p)) +
geom_bar(stat = "identity") +
geom_errorbar(aes(ymin=response-SE, ymax=response+SE), width=.2,
position=position_dodge(0.06)) +
coord_flip() +
ylim(0,6) +
labs(y="n tests with high SCC",title="Streptococcus dysgalactiae", fill="IMI type") +
scale_x_discrete(limits = rev) +
theme_classic() +
theme(plot.title=element_text(face="italic"),axis.text.y=element_blank(),axis.title.y = element_blank()) +
scale_fill_manual(labels = c("No IMI", "Transient IMI", "Persistent IMI"),
values = c("#009999", "#FF9944","#990000"))
contrasts_p8 <- as.data.frame(exp(Confint(nhscc_8_adj)))
contrasts_p8$comparison <- row.names(contrasts_p8)
contrasts_p8 <- contrasts_p8 %>% filter(grepl("IMI.cow.p",comparison))
contrasts_p8 <- contrasts_p8 %>% rename("lower"=c(2),"upper"=c(3))
# Create a new data frame with the extra observation
extra_observation <- data.frame(comparison = "0",
Estimate = 1,
upper = 1,
lower = 1)
# Combine the extra observation with the original data frame
contrasts_p8_with_extra <- rbind(contrasts_p8, extra_observation) %>% arrange(comparison)
contrasts_p8_with_extra$strep.dys.IMI.cow <- as.factor(row_number(contrasts_p8_with_extra))
nhscc_plot_8_b <- ggplot(contrasts_p8_with_extra, aes(x=strep.dys.IMI.cow,y=Estimate,colour=strep.dys.IMI.cow)) +
geom_point() +
geom_errorbar(aes(ymin=lower, ymax=upper), width=.2,
position=position_dodge(0.05)) +
ylim(0,5) +
coord_flip() +
scale_x_discrete(limits = rev) +
geom_hline(yintercept=1, linetype="dashed", size=0.5) +
theme_classic() +
theme(axis.text.y=element_blank(),axis.title.y = element_blank()) +
scale_colour_manual(labels = c("No IMI", "Transient IMI", "Persistent IMI"),
values = c("#009999", "#FF9933","#990000"))
# ---- Strep sp ----
nhscc_9 <- glmer.nb(n_high_scc~as.factor(strep.sp.IMI.cow.p)+calving.sau+calving.nas+(1|Site),data=df_dhi_cow)
summary(nhscc_9)## Generalized linear mixed model fit by maximum likelihood (Laplace
## Approximation) [glmerMod]
## Family: Negative Binomial(1.7424) ( log )
## Formula:
## n_high_scc ~ as.factor(strep.sp.IMI.cow.p) + calving.sau + calving.nas +
## (1 | Site)
## Data: df_dhi_cow
##
## AIC BIC logLik deviance df.resid
## 1025.4 1052.1 -505.7 1011.4 326
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -1.0817 -0.6890 -0.4988 0.3713 5.6669
##
## Random effects:
## Groups Name Variance Std.Dev.
## Site (Intercept) 0.2259 0.4753
## Number of obs: 333, groups: Site, 4
##
## Fixed effects:
## Estimate Std. Error z value Pr(>|z|)
## (Intercept) -0.09497 0.25780 -0.368 0.713
## as.factor(strep.sp.IMI.cow.p)1 0.29442 0.23997 1.227 0.220
## as.factor(strep.sp.IMI.cow.p)2 0.54983 0.34174 1.609 0.108
## calving.sau1 0.92939 0.15586 5.963 2.48e-09 ***
## calving.nas1 0.13986 0.14992 0.933 0.351
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## (Intr) a.(..IMI..)1 a.(..IMI..)2 clvng.s1
## a.(..IMI..)1 -0.092
## a.(..IMI..)2 -0.042 0.065
## calving.sa1 -0.187 0.027 0.060
## calving.ns1 -0.205 0.125 -0.053 0.227emmeans(nhscc_9,revpairwise~strep.sp.IMI.cow.p,type="response")## $emmeans
## strep.sp.IMI.cow.p response SE df asymp.LCL asymp.UCL
## 0 1.55 0.397 Inf 0.94 2.56
## 1 2.08 0.713 Inf 1.07 4.07
## 2 2.69 1.126 Inf 1.18 6.11
##
## Results are averaged over the levels of: calving.sau, calving.nas
## Confidence level used: 0.95
## Intervals are back-transformed from the log scale
##
## $contrasts
## contrast ratio SE df null z.ratio p.value
## strep.sp.IMI.cow.p1 / strep.sp.IMI.cow.p0 1.34 0.322 Inf 1 1.227 0.4372
## strep.sp.IMI.cow.p2 / strep.sp.IMI.cow.p0 1.73 0.592 Inf 1 1.609 0.2418
## strep.sp.IMI.cow.p2 / strep.sp.IMI.cow.p1 1.29 0.522 Inf 1 0.631 0.8029
##
## Results are averaged over the levels of: calving.sau, calving.nas
## P value adjustment: tukey method for comparing a family of 3 estimates
## Tests are performed on the log scalenhscc_9_adj <- glmer.nb(n_high_scc~as.factor(strep.sp.IMI.cow.p)+calving.sau+calving.nas+(1|Site),data=df_dhi_cow)
summary(nhscc_9_adj)## Generalized linear mixed model fit by maximum likelihood (Laplace
## Approximation) [glmerMod]
## Family: Negative Binomial(1.7424) ( log )
## Formula:
## n_high_scc ~ as.factor(strep.sp.IMI.cow.p) + calving.sau + calving.nas +
## (1 | Site)
## Data: df_dhi_cow
##
## AIC BIC logLik deviance df.resid
## 1025.4 1052.1 -505.7 1011.4 326
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -1.0817 -0.6890 -0.4988 0.3713 5.6669
##
## Random effects:
## Groups Name Variance Std.Dev.
## Site (Intercept) 0.2259 0.4753
## Number of obs: 333, groups: Site, 4
##
## Fixed effects:
## Estimate Std. Error z value Pr(>|z|)
## (Intercept) -0.09497 0.25780 -0.368 0.713
## as.factor(strep.sp.IMI.cow.p)1 0.29442 0.23997 1.227 0.220
## as.factor(strep.sp.IMI.cow.p)2 0.54983 0.34174 1.609 0.108
## calving.sau1 0.92939 0.15586 5.963 2.48e-09 ***
## calving.nas1 0.13986 0.14992 0.933 0.351
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## (Intr) a.(..IMI..)1 a.(..IMI..)2 clvng.s1
## a.(..IMI..)1 -0.092
## a.(..IMI..)2 -0.042 0.065
## calving.sa1 -0.187 0.027 0.060
## calving.ns1 -0.205 0.125 -0.053 0.227exp(Confint(nhscc_9_adj))## Estimate 2.5 % 97.5 %
## (Intercept) 0.9094011 0.5486801 1.507272
## as.factor(strep.sp.IMI.cow.p)1 1.3423495 0.8386923 2.148466
## as.factor(strep.sp.IMI.cow.p)2 1.7329525 0.8869519 3.385893
## calving.sau1 2.5329572 1.8661945 3.437944
## calving.nas1 1.1501112 0.8572821 1.542964emmeans_p9 <- emmeans(nhscc_9_adj,revpairwise~strep.sp.IMI.cow.p,type="response")
performance::icc(nhscc_9_adj)## Warning: mu of 1.2 is too close to zero, estimate of random effect variances may
## be unreliable.## # Intraclass Correlation Coefficient
##
## Adjusted ICC: 0.206
## Unadjusted ICC: 0.182emmeans_p9## $emmeans
## strep.sp.IMI.cow.p response SE df asymp.LCL asymp.UCL
## 0 1.55 0.397 Inf 0.94 2.56
## 1 2.08 0.713 Inf 1.07 4.07
## 2 2.69 1.126 Inf 1.18 6.11
##
## Results are averaged over the levels of: calving.sau, calving.nas
## Confidence level used: 0.95
## Intervals are back-transformed from the log scale
##
## $contrasts
## contrast ratio SE df null z.ratio p.value
## strep.sp.IMI.cow.p1 / strep.sp.IMI.cow.p0 1.34 0.322 Inf 1 1.227 0.4372
## strep.sp.IMI.cow.p2 / strep.sp.IMI.cow.p0 1.73 0.592 Inf 1 1.609 0.2418
## strep.sp.IMI.cow.p2 / strep.sp.IMI.cow.p1 1.29 0.522 Inf 1 0.631 0.8029
##
## Results are averaged over the levels of: calving.sau, calving.nas
## P value adjustment: tukey method for comparing a family of 3 estimates
## Tests are performed on the log scaleadj_means_p9 <- as.data.frame(emmeans_p9$emmeans)
adj_means_p9## strep.sp.IMI.cow.p response SE df asymp.LCL asymp.UCL
## 0 1.552170 0.3968246 Inf 0.9404222 2.561862
## 1 2.083555 0.7127592 Inf 1.0656608 4.073716
## 2 2.689837 1.1257382 Inf 1.1843633 6.108955
##
## Results are averaged over the levels of: calving.sau, calving.nas
## Confidence level used: 0.95
## Intervals are back-transformed from the log scalenhscc_plot_9 <- ggplot(adj_means_p9, aes(x=strep.sp.IMI.cow.p, y=response, fill=strep.sp.IMI.cow.p)) +
geom_bar(stat = "identity") +
geom_errorbar(aes(ymin=response-SE, ymax=response+SE), width=.2,
position=position_dodge(0.06)) +
coord_flip() +
ylim(0,6) +
labs(y="n tests with high SCC",title=strep_sp_title, fill="IMI type") +
scale_x_discrete(limits = rev) +
theme_classic() +
theme(axis.text.y=element_blank(),axis.title.y = element_blank()) +
scale_fill_manual(labels = c("No IMI", "Transient IMI", "Persistent IMI"),
values = c("#009999", "#FF9944","#990000"))
# IRR and 95% CI
contrasts_p9 <- as.data.frame(exp(Confint(nhscc_9_adj)))
contrasts_p9$comparison <- row.names(contrasts_p9)
contrasts_p9 <- contrasts_p9 %>% filter(grepl("IMI.cow.p",comparison))
contrasts_p9 <- contrasts_p9 %>% rename("lower"=c(2),"upper"=c(3))
# Create a new data frame with the extra observation
extra_observation <- data.frame(comparison = "0",
Estimate = 1,
upper = 1,
lower = 1)
# Combine the extra observation with the original data frame
contrasts_p9_with_extra <- rbind(contrasts_p9, extra_observation) %>% arrange(comparison)
contrasts_p9_with_extra$strep.sp.IMI.cow <- as.factor(row_number(contrasts_p9_with_extra))
nhscc_plot_9_b <- ggplot(contrasts_p9_with_extra, aes(x=strep.sp.IMI.cow,y=Estimate,colour=strep.sp.IMI.cow)) +
geom_point() +
geom_errorbar(aes(ymin=lower, ymax=upper), width=.2,
position=position_dodge(0.05)) +
ylim(0,5) +
coord_flip() +
scale_x_discrete(limits = rev) +
geom_hline(yintercept=1, linetype="dashed", size=0.5) +
theme_classic() +
theme(axis.text.y=element_blank(),axis.title.y = element_blank()) +
scale_colour_manual(labels = c("No IMI", "Transient IMI", "Persistent IMI"),
values = c("#009999", "#FF9933","#990000"))
# ---- Strep like -----
nhscc_10 <- glmer.nb(n_high_scc~as.factor(strep.like.IMI.cow.p)+calving.sau+calving.nas+(1|Site),data=df_dhi_cow)
summary(nhscc_10)## Generalized linear mixed model fit by maximum likelihood (Laplace
## Approximation) [glmerMod]
## Family: Negative Binomial(1.6721) ( log )
## Formula: n_high_scc ~ as.factor(strep.like.IMI.cow.p) + calving.sau +
## calving.nas + (1 | Site)
## Data: df_dhi_cow
##
## AIC BIC logLik deviance df.resid
## 1028.2 1054.9 -507.1 1014.2 326
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -1.0498 -0.6862 -0.5328 0.4150 5.5629
##
## Random effects:
## Groups Name Variance Std.Dev.
## Site (Intercept) 0.25 0.5
## Number of obs: 333, groups: Site, 4
##
## Fixed effects:
## Estimate Std. Error z value Pr(>|z|)
## (Intercept) -0.05781 0.27094 -0.213 0.831
## as.factor(strep.like.IMI.cow.p)1 -0.05842 0.21844 -0.267 0.789
## as.factor(strep.like.IMI.cow.p)2 0.41573 0.43166 0.963 0.336
## calving.sau1 0.92630 0.15795 5.865 4.5e-09 ***
## calving.nas1 0.13653 0.14990 0.911 0.362
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## (Intr) a.(..IMI..)1 a.(..IMI..)2 clvng.s1
## a.(..IMI..)1 -0.128
## a.(..IMI..)2 -0.070 0.108
## calving.sa1 -0.185 0.042 0.091
## calving.ns1 -0.195 0.029 0.036 0.233emmeans(nhscc_10,revpairwise~strep.like.IMI.cow.p,type="response")## $emmeans
## strep.like.IMI.cow.p response SE df asymp.LCL asymp.UCL
## 0 1.61 0.431 Inf 0.948 2.72
## 1 1.51 0.496 Inf 0.797 2.88
## 2 2.43 1.219 Inf 0.912 6.49
##
## Results are averaged over the levels of: calving.sau, calving.nas
## Confidence level used: 0.95
## Intervals are back-transformed from the log scale
##
## $contrasts
## contrast ratio SE df null z.ratio
## strep.like.IMI.cow.p1 / strep.like.IMI.cow.p0 0.943 0.206 Inf 1 -0.267
## strep.like.IMI.cow.p2 / strep.like.IMI.cow.p0 1.515 0.654 Inf 1 0.963
## strep.like.IMI.cow.p2 / strep.like.IMI.cow.p1 1.607 0.743 Inf 1 1.026
## p.value
## 0.9613
## 0.6003
## 0.5605
##
## Results are averaged over the levels of: calving.sau, calving.nas
## P value adjustment: tukey method for comparing a family of 3 estimates
## Tests are performed on the log scalenhscc_10_adj <- glmer.nb(n_high_scc~as.factor(strep.like.IMI.cow.p)+calving.sau+calving.nas+(1|Site),data=df_dhi_cow)
summary(nhscc_10_adj)## Generalized linear mixed model fit by maximum likelihood (Laplace
## Approximation) [glmerMod]
## Family: Negative Binomial(1.6721) ( log )
## Formula: n_high_scc ~ as.factor(strep.like.IMI.cow.p) + calving.sau +
## calving.nas + (1 | Site)
## Data: df_dhi_cow
##
## AIC BIC logLik deviance df.resid
## 1028.2 1054.9 -507.1 1014.2 326
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -1.0498 -0.6862 -0.5328 0.4150 5.5629
##
## Random effects:
## Groups Name Variance Std.Dev.
## Site (Intercept) 0.25 0.5
## Number of obs: 333, groups: Site, 4
##
## Fixed effects:
## Estimate Std. Error z value Pr(>|z|)
## (Intercept) -0.05781 0.27094 -0.213 0.831
## as.factor(strep.like.IMI.cow.p)1 -0.05842 0.21844 -0.267 0.789
## as.factor(strep.like.IMI.cow.p)2 0.41573 0.43166 0.963 0.336
## calving.sau1 0.92630 0.15795 5.865 4.5e-09 ***
## calving.nas1 0.13653 0.14990 0.911 0.362
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## (Intr) a.(..IMI..)1 a.(..IMI..)2 clvng.s1
## a.(..IMI..)1 -0.128
## a.(..IMI..)2 -0.070 0.108
## calving.sa1 -0.185 0.042 0.091
## calving.ns1 -0.195 0.029 0.036 0.233exp(Confint(nhscc_10_adj))## Estimate 2.5 % 97.5 %
## (Intercept) 0.9438254 0.5549652 1.605157
## as.factor(strep.like.IMI.cow.p)1 0.9432519 0.6147397 1.447318
## as.factor(strep.like.IMI.cow.p)2 1.5154751 0.6503038 3.531680
## calving.sau1 2.5251417 1.8528436 3.441381
## calving.nas1 1.1462949 0.8544859 1.537757emmeans(nhscc_10_adj,revpairwise~strep.like.IMI.cow.p,type="response")## $emmeans
## strep.like.IMI.cow.p response SE df asymp.LCL asymp.UCL
## 0 1.61 0.431 Inf 0.948 2.72
## 1 1.51 0.496 Inf 0.797 2.88
## 2 2.43 1.219 Inf 0.912 6.49
##
## Results are averaged over the levels of: calving.sau, calving.nas
## Confidence level used: 0.95
## Intervals are back-transformed from the log scale
##
## $contrasts
## contrast ratio SE df null z.ratio
## strep.like.IMI.cow.p1 / strep.like.IMI.cow.p0 0.943 0.206 Inf 1 -0.267
## strep.like.IMI.cow.p2 / strep.like.IMI.cow.p0 1.515 0.654 Inf 1 0.963
## strep.like.IMI.cow.p2 / strep.like.IMI.cow.p1 1.607 0.743 Inf 1 1.026
## p.value
## 0.9613
## 0.6003
## 0.5605
##
## Results are averaged over the levels of: calving.sau, calving.nas
## P value adjustment: tukey method for comparing a family of 3 estimates
## Tests are performed on the log scaleemmeans_p10 <- emmeans(nhscc_10_adj,revpairwise~strep.like.IMI.cow.p,type="response")
performance::icc(nhscc_10_adj)## Warning: mu of 1.2 is too close to zero, estimate of random effect variances may
## be unreliable.## # Intraclass Correlation Coefficient
##
## Adjusted ICC: 0.221
## Unadjusted ICC: 0.199emmeans_p10## $emmeans
## strep.like.IMI.cow.p response SE df asymp.LCL asymp.UCL
## 0 1.61 0.431 Inf 0.948 2.72
## 1 1.51 0.496 Inf 0.797 2.88
## 2 2.43 1.219 Inf 0.912 6.49
##
## Results are averaged over the levels of: calving.sau, calving.nas
## Confidence level used: 0.95
## Intervals are back-transformed from the log scale
##
## $contrasts
## contrast ratio SE df null z.ratio
## strep.like.IMI.cow.p1 / strep.like.IMI.cow.p0 0.943 0.206 Inf 1 -0.267
## strep.like.IMI.cow.p2 / strep.like.IMI.cow.p0 1.515 0.654 Inf 1 0.963
## strep.like.IMI.cow.p2 / strep.like.IMI.cow.p1 1.607 0.743 Inf 1 1.026
## p.value
## 0.9613
## 0.6003
## 0.5605
##
## Results are averaged over the levels of: calving.sau, calving.nas
## P value adjustment: tukey method for comparing a family of 3 estimates
## Tests are performed on the log scaleadj_means_p10 <- as.data.frame(emmeans_p10$emmeans)
adj_means_p10## strep.like.IMI.cow.p response SE df asymp.LCL asymp.UCL
## 0 1.605768 0.4314697 Inf 0.9483418 2.718946
## 1 1.514643 0.4959215 Inf 0.7972815 2.877459
## 2 2.433501 1.2185379 Inf 0.9120224 6.493181
##
## Results are averaged over the levels of: calving.sau, calving.nas
## Confidence level used: 0.95
## Intervals are back-transformed from the log scalenhscc_plot_10 <- ggplot(adj_means_p10, aes(x=strep.like.IMI.cow.p, y=response, fill=strep.like.IMI.cow.p)) +
geom_bar(stat = "identity") +
geom_errorbar(aes(ymin=response-SE, ymax=response+SE), width=.2,
position=position_dodge(0.06)) +
coord_flip() +
ylim(0,6) +
labs(y="n tests with high SCC",title=strep_like_title, fill="IMI type") +
scale_x_discrete(limits = rev) +
theme_classic() +
theme(axis.text.y=element_blank(),axis.title.y = element_blank()) +
scale_fill_manual(labels = c("No IMI", "Transient IMI", "Persistent IMI"),
values = c("#009999", "#FF9944","#990000"))
# IRR and 95% CI
contrasts_p10 <- as.data.frame(exp(Confint(nhscc_10_adj)))
contrasts_p10$comparison <- row.names(contrasts_p10)
contrasts_p10 <- contrasts_p10 %>% filter(grepl("IMI.cow.p",comparison))
contrasts_p10 <- contrasts_p10 %>% rename("lower"=c(2),"upper"=c(3))
# Create a new data frame with the extra observation
extra_observation <- data.frame(comparison = "0",
Estimate = 1,
upper = 1,
lower = 1)
# Combine the extra observation with the original data frame
contrasts_p10_with_extra <- rbind(contrasts_p10, extra_observation) %>% arrange(comparison)
contrasts_p10_with_extra$strep.like.IMI.cow <- contrasts_p10_with_extra$comparison
nhscc_plot_10_b <- ggplot(contrasts_p10_with_extra, aes(x=strep.like.IMI.cow,y=Estimate,colour=strep.like.IMI.cow)) +
geom_point() +
geom_errorbar(aes(ymin=lower, ymax=upper), width=.2,
position=position_dodge(0.05)) +
ylim(0,5) +
coord_flip() +
scale_x_discrete(limits = rev) +
geom_hline(yintercept=1, linetype="dashed", size=0.5) +
theme_classic() +
theme(axis.text.y=element_blank(),axis.title.y = element_blank()) +
scale_colour_manual(labels = c("No IMI", "Transient IMI", "Persistent IMI"),
values = c("#009999", "#FF9933","#990000"))
# ---- Enterococcus -----
nhscc_11 <- glmer.nb(n_high_scc~as.factor(enterococcus.spp.IMI.cow.p)+calving.sau+calving.nas+(1|Site),data=df_dhi_cow)
summary(nhscc_11)## Generalized linear mixed model fit by maximum likelihood (Laplace
## Approximation) [glmerMod]
## Family: Negative Binomial(1.7715) ( log )
## Formula: n_high_scc ~ as.factor(enterococcus.spp.IMI.cow.p) + calving.sau +
## calving.nas + (1 | Site)
## Data: df_dhi_cow
##
## AIC BIC logLik deviance df.resid
## 1022.6 1049.3 -504.3 1008.6 326
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -1.0663 -0.6879 -0.5381 0.3775 6.2141
##
## Random effects:
## Groups Name Variance Std.Dev.
## Site (Intercept) 0.2878 0.5365
## Number of obs: 333, groups: Site, 4
##
## Fixed effects:
## Estimate Std. Error z value Pr(>|z|)
## (Intercept) -0.09739 0.28684 -0.340 0.7342
## as.factor(enterococcus.spp.IMI.cow.p)1 0.20441 0.27569 0.741 0.4584
## as.factor(enterococcus.spp.IMI.cow.p)2 1.31320 0.51812 2.535 0.0113 *
## calving.sau1 0.92967 0.15491 6.001 1.96e-09 ***
## calving.nas1 0.12893 0.14871 0.867 0.3859
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## (Intr) a.(..IMI..)1 a.(..IMI..)2 clvng.s1
## a.(..IMI..)1 -0.105
## a.(..IMI..)2 -0.051 0.129
## calving.sa1 -0.166 0.004 0.044
## calving.ns1 -0.182 0.062 -0.002 0.228emmeans(nhscc_11,revpairwise~enterococcus.spp.IMI.cow.p,type="response")## $emmeans
## enterococcus.spp.IMI.cow.p response SE df asymp.LCL asymp.UCL
## 0 1.54 0.439 Inf 0.881 2.69
## 1 1.89 0.716 Inf 0.899 3.97
## 2 5.73 3.329 Inf 1.833 17.89
##
## Results are averaged over the levels of: calving.sau, calving.nas
## Confidence level used: 0.95
## Intervals are back-transformed from the log scale
##
## $contrasts
## contrast ratio SE df null
## enterococcus.spp.IMI.cow.p1 / enterococcus.spp.IMI.cow.p0 1.23 0.338 Inf 1
## enterococcus.spp.IMI.cow.p2 / enterococcus.spp.IMI.cow.p0 3.72 1.926 Inf 1
## enterococcus.spp.IMI.cow.p2 / enterococcus.spp.IMI.cow.p1 3.03 1.681 Inf 1
## z.ratio p.value
## 0.741 0.7388
## 2.535 0.0303
## 2.000 0.1123
##
## Results are averaged over the levels of: calving.sau, calving.nas
## P value adjustment: tukey method for comparing a family of 3 estimates
## Tests are performed on the log scalenhscc_11_adj <- glmer.nb(n_high_scc~as.factor(enterococcus.spp.IMI.cow.p)+calving.sau+calving.nas+(1|Site),data=df_dhi_cow)
summary(nhscc_11_adj)## Generalized linear mixed model fit by maximum likelihood (Laplace
## Approximation) [glmerMod]
## Family: Negative Binomial(1.7715) ( log )
## Formula: n_high_scc ~ as.factor(enterococcus.spp.IMI.cow.p) + calving.sau +
## calving.nas + (1 | Site)
## Data: df_dhi_cow
##
## AIC BIC logLik deviance df.resid
## 1022.6 1049.3 -504.3 1008.6 326
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -1.0663 -0.6879 -0.5381 0.3775 6.2141
##
## Random effects:
## Groups Name Variance Std.Dev.
## Site (Intercept) 0.2878 0.5365
## Number of obs: 333, groups: Site, 4
##
## Fixed effects:
## Estimate Std. Error z value Pr(>|z|)
## (Intercept) -0.09739 0.28684 -0.340 0.7342
## as.factor(enterococcus.spp.IMI.cow.p)1 0.20441 0.27569 0.741 0.4584
## as.factor(enterococcus.spp.IMI.cow.p)2 1.31320 0.51812 2.535 0.0113 *
## calving.sau1 0.92967 0.15491 6.001 1.96e-09 ***
## calving.nas1 0.12893 0.14871 0.867 0.3859
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## (Intr) a.(..IMI..)1 a.(..IMI..)2 clvng.s1
## a.(..IMI..)1 -0.105
## a.(..IMI..)2 -0.051 0.129
## calving.sa1 -0.166 0.004 0.044
## calving.ns1 -0.182 0.062 -0.002 0.228exp(Confint(nhscc_11_adj))## Estimate 2.5 % 97.5 %
## (Intercept) 0.9072028 0.5170648 1.591709
## as.factor(enterococcus.spp.IMI.cow.p)1 1.2268053 0.7146753 2.105923
## as.factor(enterococcus.spp.IMI.cow.p)2 3.7180497 1.3467669 10.264504
## calving.sau1 2.5336638 1.8702195 3.432459
## calving.nas1 1.1376116 0.8499899 1.522560emmeans_p11 <- emmeans(nhscc_11_adj,revpairwise~enterococcus.spp.IMI.cow.p,type="response")
performance::icc(nhscc_11_adj)## Warning: mu of 1.2 is too close to zero, estimate of random effect variances may
## be unreliable.## # Intraclass Correlation Coefficient
##
## Adjusted ICC: 0.250
## Unadjusted ICC: 0.222emmeans_p11## $emmeans
## enterococcus.spp.IMI.cow.p response SE df asymp.LCL asymp.UCL
## 0 1.54 0.439 Inf 0.881 2.69
## 1 1.89 0.716 Inf 0.899 3.97
## 2 5.73 3.329 Inf 1.833 17.89
##
## Results are averaged over the levels of: calving.sau, calving.nas
## Confidence level used: 0.95
## Intervals are back-transformed from the log scale
##
## $contrasts
## contrast ratio SE df null
## enterococcus.spp.IMI.cow.p1 / enterococcus.spp.IMI.cow.p0 1.23 0.338 Inf 1
## enterococcus.spp.IMI.cow.p2 / enterococcus.spp.IMI.cow.p0 3.72 1.926 Inf 1
## enterococcus.spp.IMI.cow.p2 / enterococcus.spp.IMI.cow.p1 3.03 1.681 Inf 1
## z.ratio p.value
## 0.741 0.7388
## 2.535 0.0303
## 2.000 0.1123
##
## Results are averaged over the levels of: calving.sau, calving.nas
## P value adjustment: tukey method for comparing a family of 3 estimates
## Tests are performed on the log scaleadj_means_p11 <- as.data.frame(emmeans_p11$emmeans)
adj_means_p11## enterococcus.spp.IMI.cow.p response SE df asymp.LCL asymp.UCL
## 0 1.540196 0.439083 Inf 0.8808725 2.693015
## 1 1.889520 0.715580 Inf 0.8994828 3.969266
## 2 5.726524 3.328765 Inf 1.8327237 17.893082
##
## Results are averaged over the levels of: calving.sau, calving.nas
## Confidence level used: 0.95
## Intervals are back-transformed from the log scaleadj_means_p11$upper_SE <- adj_means_p11$response + adj_means_p11$SE
adj_means_p11$upper_SE <- ifelse(adj_means_p11$upper_SE>6,6,adj_means_p11$upper_SE)
adj_means_p11$enterococcus.spp.IMI.cow.p <- as.factor(adj_means_p11$enterococcus.spp.IMI.cow.p)
nhscc_plot_11 <- ggplot(adj_means_p11, aes(x=enterococcus.spp.IMI.cow.p, y=response, fill=enterococcus.spp.IMI.cow.p)) +
geom_bar(stat = "identity") +
geom_errorbar(aes(ymin=response-SE, ymax=upper_SE), width=.2,
position=position_dodge(0.06)) +
coord_flip() +
ylim(0,6) +
labs(y="n tests with high SCC",title=enterococcus_genus_title, fill="IMI type") +
scale_x_discrete(limits = rev) +
theme_classic() +
theme(axis.text.y=element_blank(),axis.title.y = element_blank()) +
scale_fill_manual(labels = c("No IMI", "Transient IMI", "Persistent IMI"),
values = c("#009999", "#FF9944","#990000"))
# IRR and 95% CI
contrasts_p11 <- as.data.frame(exp(Confint(nhscc_11_adj)))
contrasts_p11$comparison <- row.names(contrasts_p11)
contrasts_p11 <- contrasts_p11 %>% filter(grepl("IMI.cow.p",comparison))
contrasts_p11 <- contrasts_p11 %>% rename("lower"=c(2),"upper"=c(3))
# Create a new data frame with the extra observation
extra_observation <- data.frame(comparison = "0",
Estimate = 1,
upper = 1,
lower = 1)
# Combine the extra observation with the original data frame
contrasts_p11_with_extra <- rbind(contrasts_p11, extra_observation) %>% arrange(comparison)
contrasts_p11_with_extra$upper <- ifelse(contrasts_p11_with_extra$upper>5,5,contrasts_p11_with_extra$upper)
contrasts_p11_with_extra$entero.IMI.cow <- contrasts_p11_with_extra$comparison
nhscc_plot_11_b <- ggplot(contrasts_p11_with_extra, aes(x=entero.IMI.cow,y=Estimate,colour=entero.IMI.cow)) +
geom_point() +
geom_errorbar(aes(ymin=lower, ymax=upper), width=.2,
position=position_dodge(0.05)) +
ylim(0,5) +
coord_flip() +
scale_x_discrete(limits = rev) +
geom_hline(yintercept=1, linetype="dashed", size=0.5) +
theme_classic() +
theme(axis.text.y=element_blank(),axis.title.y = element_blank()) +
scale_colour_manual(labels = c("No IMI", "Transient IMI", "Persistent IMI"),
values = c("#009999", "#FF9933","#990000"))
# ---- gram-negative ----
nhscc_12 <- glmer.nb(n_high_scc~as.factor(gram.neg.IMI.cow)+calving.sau.sslo+calving.nas+(1|Site),data=df_dhi_cow)
summary(nhscc_12)## Generalized linear mixed model fit by maximum likelihood (Laplace
## Approximation) [glmerMod]
## Family: Negative Binomial(1.7309) ( log )
## Formula: n_high_scc ~ as.factor(gram.neg.IMI.cow) + calving.sau.sslo +
## calving.nas + (1 | Site)
## Data: df_dhi_cow
##
## AIC BIC logLik deviance df.resid
## 1018.9 1041.8 -503.5 1006.9 327
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -1.1164 -0.6677 -0.5929 0.4443 6.2294
##
## Random effects:
## Groups Name Variance Std.Dev.
## Site (Intercept) 0.201 0.4483
## Number of obs: 333, groups: Site, 4
##
## Fixed effects:
## Estimate Std. Error z value Pr(>|z|)
## (Intercept) -0.25356 0.25192 -1.007 0.314
## as.factor(gram.neg.IMI.cow)1 0.00776 0.25805 0.030 0.976
## calving.sau.sslo1 0.95342 0.14538 6.558 5.44e-11 ***
## calving.nas1 0.19609 0.15166 1.293 0.196
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## (Intr) a.(..I clv..1
## a.(..IMI.)1 -0.075
## clvng.s.ss1 -0.301 0.009
## calving.ns1 -0.236 -0.053 0.281emmeans(nhscc_12,revpairwise~gram.neg.IMI.cow,type="response")## $emmeans
## gram.neg.IMI.cow response SE df asymp.LCL asymp.UCL
## 0 1.38 0.331 Inf 0.862 2.21
## 1 1.39 0.466 Inf 0.720 2.68
##
## Results are averaged over the levels of: calving.sau.sslo, calving.nas
## Confidence level used: 0.95
## Intervals are back-transformed from the log scale
##
## $contrasts
## contrast ratio SE df null z.ratio p.value
## gram.neg.IMI.cow1 / gram.neg.IMI.cow0 1.01 0.26 Inf 1 0.030 0.9760
##
## Results are averaged over the levels of: calving.sau.sslo, calving.nas
## Tests are performed on the log scalenhscc_12_adj <- glmer.nb(n_high_scc~as.factor(gram.neg.IMI.cow)+calving.sau.sslo+calving.nas+(1|Site),data=df_dhi_cow,
family=poisson(link="log"))
summary(nhscc_12_adj)## Generalized linear mixed model fit by maximum likelihood (Laplace
## Approximation) [glmerMod]
## Family: Negative Binomial(1.7309) ( log )
## Formula: n_high_scc ~ as.factor(gram.neg.IMI.cow) + calving.sau.sslo +
## calving.nas + (1 | Site)
## Data: df_dhi_cow
##
## AIC BIC logLik deviance df.resid
## 1018.9 1041.8 -503.5 1006.9 327
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -1.1164 -0.6677 -0.5929 0.4443 6.2294
##
## Random effects:
## Groups Name Variance Std.Dev.
## Site (Intercept) 0.201 0.4483
## Number of obs: 333, groups: Site, 4
##
## Fixed effects:
## Estimate Std. Error z value Pr(>|z|)
## (Intercept) -0.25356 0.25192 -1.007 0.314
## as.factor(gram.neg.IMI.cow)1 0.00776 0.25805 0.030 0.976
## calving.sau.sslo1 0.95342 0.14538 6.558 5.44e-11 ***
## calving.nas1 0.19609 0.15166 1.293 0.196
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## (Intr) a.(..I clv..1
## a.(..IMI.)1 -0.075
## clvng.s.ss1 -0.301 0.009
## calving.ns1 -0.236 -0.053 0.281exp(Confint(nhscc_12_adj))## Estimate 2.5 % 97.5 %
## (Intercept) 0.7760319 0.4736407 1.271482
## as.factor(gram.neg.IMI.cow)1 1.0077903 0.6077429 1.671169
## calving.sau.sslo1 2.5945692 1.9512878 3.449921
## calving.nas1 1.2166347 0.9037964 1.637758emmeans_p12 <- emmeans(nhscc_12_adj,revpairwise~gram.neg.IMI.cow,type="response")
performance::icc(nhscc_12_adj)## Warning: mu of 1.2 is too close to zero, estimate of random effect variances may
## be unreliable.## # Intraclass Correlation Coefficient
##
## Adjusted ICC: 0.188
## Unadjusted ICC: 0.157emmeans_p12## $emmeans
## gram.neg.IMI.cow response SE df asymp.LCL asymp.UCL
## 0 1.38 0.331 Inf 0.862 2.21
## 1 1.39 0.466 Inf 0.720 2.68
##
## Results are averaged over the levels of: calving.sau.sslo, calving.nas
## Confidence level used: 0.95
## Intervals are back-transformed from the log scale
##
## $contrasts
## contrast ratio SE df null z.ratio p.value
## gram.neg.IMI.cow1 / gram.neg.IMI.cow0 1.01 0.26 Inf 1 0.030 0.9760
##
## Results are averaged over the levels of: calving.sau.sslo, calving.nas
## Tests are performed on the log scaleadj_means_p12 <- as.data.frame(emmeans_p12$emmeans)
adj_means_p12## gram.neg.IMI.cow response SE df asymp.LCL asymp.UCL
## 0 1.378772 0.3307150 Inf 0.8616310 2.206294
## 1 1.389512 0.4663321 Inf 0.7197672 2.682458
##
## Results are averaged over the levels of: calving.sau.sslo, calving.nas
## Confidence level used: 0.95
## Intervals are back-transformed from the log scalenhscc_plot_12 <- ggplot(adj_means_p12, aes(x=gram.neg.IMI.cow, y=response, fill=gram.neg.IMI.cow)) +
geom_bar(stat = "identity") +
geom_errorbar(aes(ymin=response-SE, ymax=response+SE), width=.2,
position=position_dodge(0.06)) +
coord_flip() +
ylim(0,6) +
labs(y="n tests with high SCC",title="Gram-negative bacteria", fill="IMI type") +
scale_x_discrete(limits = rev) +
theme_classic() +
theme(axis.text.y=element_blank(),axis.title.y = element_blank()) +
scale_fill_manual(labels = c("No IMI", "Transient IMI", "Persistent IMI"),
values = c("#009999", "#FF9944","#990000"))
# IRR and 95% CI
contrasts_p12 <- as.data.frame(exp(Confint(nhscc_12_adj)))
contrasts_p12$comparison <- row.names(contrasts_p12)
contrasts_p12 <- contrasts_p12 %>% filter(grepl("IMI.cow",comparison))
contrasts_p12 <- contrasts_p12 %>% rename("lower"=c(2),"upper"=c(3))
# Create a new data frame with the extra observation
extra_observation <- data.frame(comparison = "0",
Estimate = 1,
upper = 1,
lower = 1)
# Combine the extra observation with the original data frame
contrasts_p12_with_extra <- rbind(contrasts_p12, extra_observation) %>% arrange(comparison)
contrasts_p12_with_extra$gram.neg.IMI.cow <- contrasts_p12_with_extra$comparison
nhscc_plot_12_b <- ggplot(contrasts_p12_with_extra, aes(x=gram.neg.IMI.cow,y=Estimate,colour=gram.neg.IMI.cow)) +
geom_point() +
geom_errorbar(aes(ymin=lower, ymax=upper), width=.2,
position=position_dodge(0.05)) +
ylim(0,5) +
coord_flip() +
scale_x_discrete(limits = rev) +
geom_hline(yintercept=1, linetype="dashed", size=0.5) +
theme_classic() +
theme(axis.text.y=element_blank(),axis.title.y = element_blank()) +
scale_colour_manual(labels = c("No IMI", "Transient IMI", "Persistent IMI"),
values = c("#009999", "#FF9933","#990000"))
# ---- Other ----
nhscc_13 <- glmer.nb(n_high_scc~as.factor(other.IMI.cow.p)+(1|Site),data=df_dhi_cow)
summary(nhscc_13)## Generalized linear mixed model fit by maximum likelihood (Laplace
## Approximation) [glmerMod]
## Family: Negative Binomial(1.3153) ( log )
## Formula: n_high_scc ~ as.factor(other.IMI.cow.p) + (1 | Site)
## Data: df_dhi_cow
##
## AIC BIC logLik deviance df.resid
## 1043.2 1062.3 -516.6 1033.2 328
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -0.9678 -0.7029 -0.3853 0.5195 4.4717
##
## Random effects:
## Groups Name Variance Std.Dev.
## Site (Intercept) 0.3325 0.5767
## Number of obs: 333, groups: Site, 4
##
## Fixed effects:
## Estimate Std. Error z value Pr(>|z|)
## (Intercept) 0.3684 0.3021 1.220 0.22263
## as.factor(other.IMI.cow.p)1 -0.4131 0.1737 -2.378 0.01739 *
## as.factor(other.IMI.cow.p)2 -1.2776 0.4038 -3.164 0.00156 **
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## (Intr) a.(.IMI..)1
## a.(.IMI..)1 -0.142
## a.(.IMI..)2 -0.082 0.205emmeans(nhscc_13,revpairwise~other.IMI.cow.p,type="response")## $emmeans
## other.IMI.cow.p response SE df asymp.LCL asymp.UCL
## 0 1.445 0.437 Inf 0.800 2.61
## 1 0.956 0.312 Inf 0.504 1.81
## 2 0.403 0.195 Inf 0.156 1.04
##
## Confidence level used: 0.95
## Intervals are back-transformed from the log scale
##
## $contrasts
## contrast ratio SE df null z.ratio p.value
## other.IMI.cow.p1 / other.IMI.cow.p0 0.662 0.115 Inf 1 -2.378 0.0458
## other.IMI.cow.p2 / other.IMI.cow.p0 0.279 0.113 Inf 1 -3.164 0.0044
## other.IMI.cow.p2 / other.IMI.cow.p1 0.421 0.171 Inf 1 -2.132 0.0835
##
## P value adjustment: tukey method for comparing a family of 3 estimates
## Tests are performed on the log scalenhscc_13_adj <- glmer.nb(n_high_scc~as.factor(other.IMI.cow.p)+calving.sau.sslo+calving.nas+(1|Site),data=df_dhi_cow)
summary(nhscc_13_adj)## Generalized linear mixed model fit by maximum likelihood (Laplace
## Approximation) [glmerMod]
## Family: Negative Binomial(1.9341) ( log )
## Formula: n_high_scc ~ as.factor(other.IMI.cow.p) + calving.sau.sslo +
## calving.nas + (1 | Site)
## Data: df_dhi_cow
##
## AIC BIC logLik deviance df.resid
## 1006.1 1032.7 -496.0 992.1 326
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -1.1616 -0.7222 -0.4305 0.4000 4.8598
##
## Random effects:
## Groups Name Variance Std.Dev.
## Site (Intercept) 0.139 0.3728
## Number of obs: 333, groups: Site, 4
##
## Fixed effects:
## Estimate Std. Error z value Pr(>|z|)
## (Intercept) -0.0626 0.2230 -0.281 0.77890
## as.factor(other.IMI.cow.p)1 -0.4231 0.1644 -2.573 0.01008 *
## as.factor(other.IMI.cow.p)2 -1.2479 0.3939 -3.168 0.00153 **
## calving.sau.sslo1 0.9399 0.1438 6.537 6.26e-11 ***
## calving.nas1 0.1213 0.1495 0.811 0.41710
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## (Intr) a.(.IMI..)1 a.(.IMI..)2 clv..1
## a.(.IMI..)1 -0.209
## a.(.IMI..)2 -0.115 0.200
## clvng.s.ss1 -0.333 -0.008 -0.001
## calving.ns1 -0.298 0.139 0.053 0.287exp(Confint(nhscc_13_adj))## Estimate 2.5 % 97.5 %
## (Intercept) 0.9393227 0.6067868 1.4540973
## as.factor(other.IMI.cow.p)1 0.6550376 0.4745873 0.9040996
## as.factor(other.IMI.cow.p)2 0.2871143 0.1326718 0.6213428
## calving.sau.sslo1 2.5596347 1.9310914 3.3927602
## calving.nas1 1.1289835 0.8422240 1.5133787emmeans_p13 <- emmeans(nhscc_13_adj,revpairwise~other.IMI.cow.p,type="response")
performance::icc(nhscc_13_adj)## Warning: mu of 1.2 is too close to zero, estimate of random effect variances may
## be unreliable.## # Intraclass Correlation Coefficient
##
## Adjusted ICC: 0.141
## Unadjusted ICC: 0.102emmeans_p13## $emmeans
## other.IMI.cow.p response SE df asymp.LCL asymp.UCL
## 0 1.597 0.331 Inf 1.064 2.40
## 1 1.046 0.252 Inf 0.653 1.68
## 2 0.458 0.195 Inf 0.199 1.06
##
## Results are averaged over the levels of: calving.sau.sslo, calving.nas
## Confidence level used: 0.95
## Intervals are back-transformed from the log scale
##
## $contrasts
## contrast ratio SE df null z.ratio p.value
## other.IMI.cow.p1 / other.IMI.cow.p0 0.655 0.108 Inf 1 -2.573 0.0272
## other.IMI.cow.p2 / other.IMI.cow.p0 0.287 0.113 Inf 1 -3.168 0.0044
## other.IMI.cow.p2 / other.IMI.cow.p1 0.438 0.173 Inf 1 -2.087 0.0925
##
## Results are averaged over the levels of: calving.sau.sslo, calving.nas
## P value adjustment: tukey method for comparing a family of 3 estimates
## Tests are performed on the log scaleadj_means_p13 <- as.data.frame(emmeans_p13$emmeans)
adj_means_p13## other.IMI.cow.p response SE df asymp.LCL asymp.UCL
## 0 1.5967892 0.330891 Inf 1.0637983 2.396823
## 1 1.0459569 0.251676 Inf 0.6526788 1.676209
## 2 0.4584611 0.195011 Inf 0.1991753 1.055284
##
## Results are averaged over the levels of: calving.sau.sslo, calving.nas
## Confidence level used: 0.95
## Intervals are back-transformed from the log scalenhscc_plot_13 <- ggplot(adj_means_p13, aes(x=other.IMI.cow.p, y=response, fill=other.IMI.cow.p)) +
geom_bar(stat = "identity") +
geom_errorbar(aes(ymin=response-SE, ymax=response+SE), width=.2,
position=position_dodge(0.06)) +
coord_flip() +
ylim(0,6) +
labs(y="n tests with high SCC",title="Others", fill="IMI type") +
scale_x_discrete(limits = rev) +
theme_classic() +
theme(axis.text.y=element_blank(),axis.title.y = element_blank()) +
scale_fill_manual(labels = c("No IMI", "Transient IMI", "Persistent IMI"),
values = c("#009999", "#FF9944","#990000"))
# IRR and 95% CI
contrasts_p13 <- as.data.frame(exp(Confint(nhscc_13_adj)))
contrasts_p13$comparison <- row.names(contrasts_p13)
contrasts_p13 <- contrasts_p13 %>% filter(grepl("IMI.cow.p",comparison))
contrasts_p13 <- contrasts_p13 %>% rename("lower"=c(2),"upper"=c(3))
# Create a new data frame with the extra observation
extra_observation <- data.frame(comparison = "0",
Estimate = 1,
upper = 1,
lower = 1)
# Combine the extra observation with the original data frame
contrasts_p13_with_extra <- rbind(contrasts_p13, extra_observation) %>% arrange(comparison)
contrasts_p13_with_extra$other.IMI.cow <- contrasts_p13_with_extra$comparison
nhscc_plot_13_b <- ggplot(contrasts_p13_with_extra, aes(x=other.IMI.cow,y=Estimate,colour=other.IMI.cow)) +
geom_point() +
geom_errorbar(aes(ymin=lower, ymax=upper), width=.2,
position=position_dodge(0.05)) +
ylim(0,5) +
coord_flip() +
scale_x_discrete(limits = rev) +
geom_hline(yintercept=1, linetype="dashed", size=0.5) +
theme_classic() +
theme(axis.text.y=element_blank(),axis.title.y = element_blank()) +
scale_colour_manual(labels = c("No IMI", "Transient IMI", "Persistent IMI"),
values = c("#009999", "#FF9933","#990000"))
# ---- Bacillus sp. ----
nhscc_14 <- glmer.nb(n_high_scc~as.factor(bacillus.IMI.cow.p)+(1|Site),data=df_dhi_cow)
summary(nhscc_14)## Generalized linear mixed model fit by maximum likelihood (Laplace
## Approximation) [glmerMod]
## Family: Negative Binomial(1.3252) ( log )
## Formula: n_high_scc ~ as.factor(bacillus.IMI.cow.p) + (1 | Site)
## Data: df_dhi_cow
##
## AIC BIC logLik deviance df.resid
## 1040.0 1059.1 -515.0 1030.0 328
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -0.9704 -0.7225 -0.3602 0.5213 5.1848
##
## Random effects:
## Groups Name Variance Std.Dev.
## Site (Intercept) 0.3631 0.6026
## Number of obs: 333, groups: Site, 4
##
## Fixed effects:
## Estimate Std. Error z value Pr(>|z|)
## (Intercept) 0.3201 0.3124 1.024 0.30561
## as.factor(bacillus.IMI.cow.p)1 -0.4844 0.2023 -2.395 0.01663 *
## as.factor(bacillus.IMI.cow.p)2 -2.1423 0.7512 -2.852 0.00435 **
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## (Intr) a.(.IMI..)1
## a.(.IMI..)1 -0.090
## a.(.IMI..)2 -0.031 0.073emmeans(nhscc_14,revpairwise~bacillus.IMI.cow.p,type="response")## $emmeans
## bacillus.IMI.cow.p response SE df asymp.LCL asymp.UCL
## 0 1.377 0.430 Inf 0.7466 2.541
## 1 0.848 0.303 Inf 0.4218 1.707
## 2 0.162 0.130 Inf 0.0334 0.782
##
## Confidence level used: 0.95
## Intervals are back-transformed from the log scale
##
## $contrasts
## contrast ratio SE df null z.ratio
## bacillus.IMI.cow.p1 / bacillus.IMI.cow.p0 0.616 0.1246 Inf 1 -2.395
## bacillus.IMI.cow.p2 / bacillus.IMI.cow.p0 0.117 0.0882 Inf 1 -2.852
## bacillus.IMI.cow.p2 / bacillus.IMI.cow.p1 0.191 0.1455 Inf 1 -2.171
## p.value
## 0.0439
## 0.0121
## 0.0762
##
## P value adjustment: tukey method for comparing a family of 3 estimates
## Tests are performed on the log scalenhscc_14_adj <- glmer.nb(n_high_scc~as.factor(bacillus.IMI.cow.p)+calving.sau.sslo+calving.nas+(1|Site),data=df_dhi_cow)
summary(nhscc_14_adj)## Generalized linear mixed model fit by maximum likelihood (Laplace
## Approximation) [glmerMod]
## Family: Negative Binomial(1.8753) ( log )
## Formula: n_high_scc ~ as.factor(bacillus.IMI.cow.p) + calving.sau.sslo +
## calving.nas + (1 | Site)
## Data: df_dhi_cow
##
## AIC BIC logLik deviance df.resid
## 1006.9 1033.6 -496.5 992.9 326
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -1.1470 -0.6874 -0.3663 0.4135 5.3495
##
## Random effects:
## Groups Name Variance Std.Dev.
## Site (Intercept) 0.1713 0.4139
## Number of obs: 333, groups: Site, 4
##
## Fixed effects:
## Estimate Std. Error z value Pr(>|z|)
## (Intercept) -0.1088 0.2384 -0.456 0.6483
## as.factor(bacillus.IMI.cow.p)1 -0.4397 0.1925 -2.284 0.0224 *
## as.factor(bacillus.IMI.cow.p)2 -1.8771 0.7436 -2.524 0.0116 *
## calving.sau.sslo1 0.8958 0.1443 6.209 5.32e-10 ***
## calving.nas1 0.1298 0.1504 0.863 0.3882
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## (Intr) a.(.IMI..)1 a.(.IMI..)2 clv..1
## a.(.IMI..)1 -0.146
## a.(.IMI..)2 -0.062 0.075
## clvng.s.ss1 -0.324 0.036 0.058
## calving.ns1 -0.269 0.120 0.037 0.291exp(Confint(nhscc_14_adj))## Estimate 2.5 % 97.5 %
## (Intercept) 0.8969462 0.56211600 1.4312214
## as.factor(bacillus.IMI.cow.p)1 0.6442200 0.44174124 0.9395080
## as.factor(bacillus.IMI.cow.p)2 0.1530353 0.03563394 0.6572333
## calving.sau.sslo1 2.4493509 1.84607725 3.2497663
## calving.nas1 1.1385588 0.84789000 1.5288729emmeans_p14 <- emmeans(nhscc_14_adj,revpairwise~bacillus.IMI.cow.p,type="response")
performance::icc(nhscc_14_adj)## Warning: mu of 1.2 is too close to zero, estimate of random effect variances may
## be unreliable.## # Intraclass Correlation Coefficient
##
## Adjusted ICC: 0.167
## Unadjusted ICC: 0.117emmeans_p14## $emmeans
## bacillus.IMI.cow.p response SE df asymp.LCL asymp.UCL
## 0 1.498 0.335 Inf 0.9660 2.32
## 1 0.965 0.270 Inf 0.5577 1.67
## 2 0.229 0.176 Inf 0.0508 1.03
##
## Results are averaged over the levels of: calving.sau.sslo, calving.nas
## Confidence level used: 0.95
## Intervals are back-transformed from the log scale
##
## $contrasts
## contrast ratio SE df null z.ratio p.value
## bacillus.IMI.cow.p1 / bacillus.IMI.cow.p0 0.644 0.124 Inf 1 -2.284 0.0580
## bacillus.IMI.cow.p2 / bacillus.IMI.cow.p0 0.153 0.114 Inf 1 -2.524 0.0311
## bacillus.IMI.cow.p2 / bacillus.IMI.cow.p1 0.238 0.179 Inf 1 -1.906 0.1368
##
## Results are averaged over the levels of: calving.sau.sslo, calving.nas
## P value adjustment: tukey method for comparing a family of 3 estimates
## Tests are performed on the log scaleadj_means_p14 <- as.data.frame(emmeans_p14$emmeans)
adj_means_p14## bacillus.IMI.cow.p response SE df asymp.LCL asymp.UCL
## 0 1.4978544 0.3351745 Inf 0.9660430 2.322431
## 1 0.9649478 0.2699302 Inf 0.5576895 1.669611
## 2 0.2292246 0.1762916 Inf 0.0507716 1.034907
##
## Results are averaged over the levels of: calving.sau.sslo, calving.nas
## Confidence level used: 0.95
## Intervals are back-transformed from the log scalenhscc_plot_14 <- ggplot(adj_means_p14, aes(x=bacillus.IMI.cow.p, y=response, fill=bacillus.IMI.cow.p)) +
geom_bar(stat = "identity") +
geom_errorbar(aes(ymin=response-SE, ymax=response+SE), width=.2,
position=position_dodge(0.06)) +
coord_flip() +
ylim(0,6) +
labs(y="n tests with high SCC",title=bacillus_title, fill="IMI type") +
scale_x_discrete(limits = rev) +
theme_classic() +
theme(axis.text.y=element_blank(),axis.title.y = element_blank()) +
scale_fill_manual(labels = c("No IMI", "Transient IMI", "Persistent IMI"),
values = c("#009999", "#FF9944","#990000"))
# IRR and 95% CI
contrasts_p14 <- as.data.frame(exp(Confint(nhscc_14_adj)))
contrasts_p14$comparison <- row.names(contrasts_p14)
contrasts_p14 <- contrasts_p14 %>% filter(grepl("IMI.cow.p",comparison))
contrasts_p14 <- contrasts_p14 %>% rename("lower"=c(2),"upper"=c(3))
# Create a new data frame with the extra observation
extra_observation <- data.frame(comparison = "0",
Estimate = 1,
upper = 1,
lower = 1)
# Combine the extra observation with the original data frame
contrasts_p14_with_extra <- rbind(contrasts_p14, extra_observation) %>% arrange(comparison)
contrasts_p14_with_extra$bacillus.IMI.cow <- contrasts_p14_with_extra$comparison
nhscc_plot_14_b <- ggplot(contrasts_p14_with_extra, aes(x=bacillus.IMI.cow,y=Estimate,colour=bacillus.IMI.cow)) +
geom_point() +
geom_errorbar(aes(ymin=lower, ymax=upper), width=.2,
position=position_dodge(0.05)) +
ylim(0,5) +
coord_flip() +
scale_x_discrete(limits = rev) +
geom_hline(yintercept=1, linetype="dashed", size=0.5) +
theme_classic() +
theme(axis.text.y=element_blank(),axis.title.y = element_blank()) +
scale_colour_manual(labels = c("No IMI", "Transient IMI", "Persistent IMI"),
values = c("#009999", "#FF9933","#990000"))
# ---- Coryne ----
nhscc_15 <- glmer.nb(n_high_scc~as.factor(coryne.IMI.cow.p)+(1|Site),data=df_dhi_cow,
family=poisson(link="log"))
summary(nhscc_15)## Generalized linear mixed model fit by maximum likelihood (Laplace
## Approximation) [glmerMod]
## Family: Negative Binomial(1.19) ( log )
## Formula: n_high_scc ~ as.factor(coryne.IMI.cow.p) + (1 | Site)
## Data: df_dhi_cow
##
## AIC BIC logLik deviance df.resid
## 1056.3 1075.3 -523.2 1046.3 328
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -0.9277 -0.6569 -0.3310 0.5641 5.1652
##
## Random effects:
## Groups Name Variance Std.Dev.
## Site (Intercept) 0.4047 0.6362
## Number of obs: 333, groups: Site, 4
##
## Fixed effects:
## Estimate Std. Error z value Pr(>|z|)
## (Intercept) 0.22488 0.32908 0.683 0.494
## as.factor(coryne.IMI.cow.p)1 -0.03028 0.25078 -0.121 0.904
## as.factor(coryne.IMI.cow.p)2 -0.48656 0.54832 -0.887 0.375
##
## Correlation of Fixed Effects:
## (Intr) a.(.IMI..)1
## a.(.IMI..)1 -0.085
## a.(.IMI..)2 -0.048 0.137emmeans(nhscc_15,revpairwise~coryne.IMI.cow.p,type="response")## $emmeans
## coryne.IMI.cow.p response SE df asymp.LCL asymp.UCL
## 0 1.25 0.412 Inf 0.657 2.39
## 1 1.21 0.482 Inf 0.558 2.64
## 2 0.77 0.482 Inf 0.226 2.62
##
## Confidence level used: 0.95
## Intervals are back-transformed from the log scale
##
## $contrasts
## contrast ratio SE df null z.ratio p.value
## coryne.IMI.cow.p1 / coryne.IMI.cow.p0 0.970 0.243 Inf 1 -0.121 0.9920
## coryne.IMI.cow.p2 / coryne.IMI.cow.p0 0.615 0.337 Inf 1 -0.887 0.6483
## coryne.IMI.cow.p2 / coryne.IMI.cow.p1 0.634 0.362 Inf 1 -0.799 0.7035
##
## P value adjustment: tukey method for comparing a family of 3 estimates
## Tests are performed on the log scalenhscc_15_adj <- glmer.nb(n_high_scc~as.factor(coryne.IMI.cow.p)+calving.sau.sslo+calving.nas+(1|Site),data=df_dhi_cow,
family=poisson(link="log"))
summary(nhscc_15_adj)## Generalized linear mixed model fit by maximum likelihood (Laplace
## Approximation) [glmerMod]
## Family: Negative Binomial(1.7459) ( log )
## Formula: n_high_scc ~ as.factor(coryne.IMI.cow.p) + calving.sau.sslo +
## calving.nas + (1 | Site)
## Data: df_dhi_cow
##
## AIC BIC logLik deviance df.resid
## 1019.5 1046.1 -502.7 1005.5 326
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -1.1191 -0.6705 -0.4961 0.4369 6.0043
##
## Random effects:
## Groups Name Variance Std.Dev.
## Site (Intercept) 0.1819 0.4265
## Number of obs: 333, groups: Site, 4
##
## Fixed effects:
## Estimate Std. Error z value Pr(>|z|)
## (Intercept) -0.23399 0.24360 -0.961 0.337
## as.factor(coryne.IMI.cow.p)1 -0.03402 0.23910 -0.142 0.887
## as.factor(coryne.IMI.cow.p)2 -0.61883 0.53583 -1.155 0.248
## calving.sau.sslo1 0.96085 0.14558 6.600 4.11e-11 ***
## calving.nas1 0.18642 0.15173 1.229 0.219
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## (Intr) a.(.IMI..)1 a.(.IMI..)2 clv..1
## a.(.IMI..)1 -0.125
## a.(.IMI..)2 -0.060 0.142
## clvng.s.ss1 -0.307 -0.010 -0.049
## calving.ns1 -0.260 0.087 0.048 0.279exp(Confint(nhscc_15_adj))## Estimate 2.5 % 97.5 %
## (Intercept) 0.7913691 0.4909348 1.275658
## as.factor(coryne.IMI.cow.p)1 0.9665524 0.6049281 1.544355
## as.factor(coryne.IMI.cow.p)2 0.5385758 0.1884283 1.539386
## calving.sau.sslo1 2.6139173 1.9650501 3.477043
## calving.nas1 1.2049248 0.8949683 1.622229emmeans_p15 <- emmeans(nhscc_15_adj,revpairwise~coryne.IMI.cow.p,type="response")
performance::icc(nhscc_15_adj)## Warning: mu of 1.2 is too close to zero, estimate of random effect variances may
## be unreliable.## # Intraclass Correlation Coefficient
##
## Adjusted ICC: 0.173
## Unadjusted ICC: 0.142emmeans_p15## $emmeans
## coryne.IMI.cow.p response SE df asymp.LCL asymp.UCL
## 0 1.404 0.323 Inf 0.894 2.21
## 1 1.357 0.426 Inf 0.734 2.51
## 2 0.756 0.431 Inf 0.248 2.31
##
## Results are averaged over the levels of: calving.sau.sslo, calving.nas
## Confidence level used: 0.95
## Intervals are back-transformed from the log scale
##
## $contrasts
## contrast ratio SE df null z.ratio p.value
## coryne.IMI.cow.p1 / coryne.IMI.cow.p0 0.967 0.231 Inf 1 -0.142 0.9889
## coryne.IMI.cow.p2 / coryne.IMI.cow.p0 0.539 0.289 Inf 1 -1.155 0.4803
## coryne.IMI.cow.p2 / coryne.IMI.cow.p1 0.557 0.309 Inf 1 -1.054 0.5427
##
## Results are averaged over the levels of: calving.sau.sslo, calving.nas
## P value adjustment: tukey method for comparing a family of 3 estimates
## Tests are performed on the log scaleadj_means_p15 <- as.data.frame(emmeans_p15$emmeans)
adj_means_p15## coryne.IMI.cow.p response SE df asymp.LCL asymp.UCL
## 0 1.4044459 0.3232936 Inf 0.8944671 2.205188
## 1 1.3574705 0.4258833 Inf 0.7339735 2.510617
## 2 0.7564005 0.4309324 Inf 0.2476338 2.310435
##
## Results are averaged over the levels of: calving.sau.sslo, calving.nas
## Confidence level used: 0.95
## Intervals are back-transformed from the log scalenhscc_plot_15 <- ggplot(adj_means_p15, aes(x=coryne.IMI.cow.p, y=response, fill=coryne.IMI.cow.p)) +
geom_bar(stat = "identity") +
geom_errorbar(aes(ymin=response-SE, ymax=response+SE), width=.2,
position=position_dodge(0.06)) +
coord_flip() +
ylim(0,6) +
labs(y="Number of DHIA tests with high SCC (SE)",title=coryne_title, fill="IMI type") +
scale_x_discrete(limits = rev) +
theme_classic() +
theme(axis.text.y=element_blank(),axis.title.y = element_blank()) +
scale_fill_manual(labels = c("No IMI", "Transient IMI", "Persistent IMI"),
values = c("#009999", "#FF9944","#990000"))
# IRR and 95% CI
contrasts_p15 <- as.data.frame(exp(Confint(nhscc_15_adj)))
contrasts_p15$comparison <- row.names(contrasts_p15)
contrasts_p15 <- contrasts_p15 %>% filter(grepl("IMI.cow.p",comparison))
contrasts_p15 <- contrasts_p15 %>% rename("lower"=c(2),"upper"=c(3))
# Create a new data frame with the extra observation
extra_observation <- data.frame(comparison = "0",
Estimate = 1,
upper = 1,
lower = 1)
# Combine the extra observation with the original data frame
contrasts_p15_with_extra <- rbind(contrasts_p15, extra_observation) %>% arrange(comparison)
contrasts_p15_with_extra$coryne.IMI.cow <- contrasts_p15_with_extra$comparison
nhscc_plot_15_b <- ggplot(contrasts_p15_with_extra, aes(x=coryne.IMI.cow,y=Estimate,colour=coryne.IMI.cow)) +
geom_point() +
geom_errorbar(aes(ymin=lower, ymax=upper), width=.2,
position=position_dodge(0.05)) +
ylim(0,5) +
coord_flip() +
labs(y="IRR (95%CI)") +
scale_x_discrete(limits = rev) +
geom_hline(yintercept=1, linetype="dashed", size=0.5) +
theme_classic() +
theme(axis.text.y=element_blank(),axis.title.y = element_blank()) +
scale_colour_manual(labels = c("No IMI", "Transient IMI", "Persistent IMI"),
values = c("#009999", "#FF9933","#990000"))
nhscc_figure_2 <- ggarrange( nhscc_plot_1 + rremove("ylab") + rremove("xlab"),
nhscc_plot_1_b + rremove("ylab") + rremove("xlab"),
nhscc_plot_2 + rremove("ylab") + rremove("xlab"),
nhscc_plot_2_b + rremove("ylab") + rremove("xlab"),
nhscc_plot_3 + rremove("ylab") + rremove("xlab"),
nhscc_plot_3_b + rremove("ylab") + rremove("xlab"),
nhscc_plot_4 + rremove("ylab") + rremove("xlab"),
nhscc_plot_4_b + rremove("ylab") + rremove("xlab"),
nhscc_plot_5 + rremove("ylab") + rremove("xlab"),
nhscc_plot_5_b + rremove("ylab") + rremove("xlab"),
nhscc_plot_6 + rremove("ylab") + rremove("xlab"),
nhscc_plot_6_b + rremove("ylab") + rremove("xlab"),
nhscc_plot_7 + rremove("ylab") + rremove("xlab"),
nhscc_plot_7_b + rremove("ylab") + rremove("xlab"),
nhscc_plot_8 + rremove("ylab") + rremove("xlab"),
nhscc_plot_8_b + rremove("ylab") + rremove("xlab"),
nhscc_plot_9 + rremove("ylab") + rremove("xlab"),
nhscc_plot_9_b + rremove("ylab") + rremove("xlab"),
nhscc_plot_10 + rremove("ylab") + rremove("xlab"),
nhscc_plot_10_b + rremove("ylab") + rremove("xlab"),
nhscc_plot_11 + rremove("ylab") + rremove("xlab"),
nhscc_plot_11_b + rremove("ylab") + rremove("xlab"),
nhscc_plot_12 + rremove("ylab") + rremove("xlab"),
nhscc_plot_12_b + rremove("ylab") + rremove("xlab"),
nhscc_plot_13 + rremove("ylab") + rremove("xlab"),
nhscc_plot_13_b + rremove("ylab") + rremove("xlab"),
nhscc_plot_14 + rremove("ylab") + rremove("xlab"),
nhscc_plot_14_b + rremove("ylab") + rremove("xlab"),
nhscc_plot_15 + rremove("ylab"),
nhscc_plot_15_b + rremove("ylab"),
ncol = 2, nrow = 15, common.legend = TRUE,align = "h")
nhscc_figure_annotated_2 <- annotate_figure(nhscc_figure_2,
left = text_grob("Taxonomic group", color = "black", rot = 90))
ggsave(plot = last_plot(),"./figures_article_2/nhscc_figure_nb.png",width = 20, height = 40, units = "cm")