-
Notifications
You must be signed in to change notification settings - Fork 3
/
forward_stepwise_regression_family.Rmd
430 lines (386 loc) · 20.7 KB
/
forward_stepwise_regression_family.Rmd
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
---
Title: "Taxa-based Models: Arcsine Square Root Transformation, Forward Stepwise Regression"
Author: Henry Paz (henry.paz@huskers.unl.edu)
Output:
html_document:
keep_md: yes
---
The following arcsine square root transforms the summarized test sets at family level and performs forward stepwise regressions of taxa-based models for average daily feed intake (ADFI), average daily gain (ADG), feed efficiency (ADG/ADFI) within heifer and steer cohort.
## Summarized test sets at family level for heifer and steer cohorts
```{r, engine='bash'}
sed 's/#OTU ID/taxonomy/g' biom_files/otu_table_test_heifer_L5.txt > r_inputs/otu_table_test_heifer_L5.txt
sed 's/#OTU ID/taxonomy/g' biom_files/otu_table_test_steer_L5.txt > r_inputs/otu_table_test_steer_L5.txt
```
```{r}
#Load packages
library(stringr)
#heifer data
heifer_test_family <- read.table("r_inputs/otu_table_test_heifer_L5.txt", sep="\t", header=T)
samples_heifer <- heifer_test_family[,-1]
taxa_heifer <- as.data.frame(str_split_fixed(heifer_test_family$taxonomy, ";", 5))
names(taxa_heifer)[5] <- "taxonomy"
taxa_heifer$taxonomy <- sub("f__", "", taxa_heifer$taxonomy)
taxa_heifer$taxonomy <- sub("\\]", "", taxa_heifer$taxonomy)
taxa_heifer$taxonomy <- sub("\\[", "", taxa_heifer$taxonomy)
taxa_heifer$taxonomy <- sub("^$", "No Assigned Family", taxa_heifer$taxonomy)
taxa_heifer$taxonomy <- sub("Other", "No Assigned Family", taxa_heifer$taxonomy)
taxa_heifer$taxonomy[1] <- "NoAssignedFamily1"
taxa_heifer$taxonomy[60] <- "NoAssignedFamily2"
taxa_heifer$taxonomy[61] <- "NoAssignedFamily3"
taxa_heifer$taxonomy[114] <- "NoAssignedFamily4"
taxa_heifer$taxonomy[134] <- "NoAssignedFamily5"
taxa_heifer$taxonomy[135] <- "NoAssignedFamily6"
taxa_heifer$taxonomy[165] <- "NoAssignedFamily7"
taxa_heifer$taxonomy[183] <- "NoAssignedFamily8"
taxa_heifer$taxonomy[188] <- "NoAssignedFamily9"
taxa_heifer$taxonomy[223] <- "NoAssignedFamily10"
taxa_heifer$taxonomy[228] <- "NoAssignedFamily11"
taxa_heifer$taxonomy[241] <- "NoAssignedFamily12"
taxa_heifer$taxonomy[264] <- "NoAssignedFamily13"
taxa_heifer$taxonomy[265] <- "NoAssignedFamily14"
taxa_heifer$taxonomy[284] <- "NoAssignedFamily15"
taxa_heifer$taxonomy[286] <- "NoAssignedFamily16"
taxa_heifer$taxonomy[289] <- "NoAssignedFamily17"
taxa_heifer$taxonomy[290] <- "NoAssignedFamily18"
taxa_heifer$taxonomy[291] <- "NoAssignedFamily19"
taxa_heifer$taxonomy[306] <- "NoAssignedFamily20"
taxa_heifer$taxonomy[167] <- "R4_45B"
#validation set
taxa_heifer$taxonomy[71] <- "S24_7"
taxa_heifer$taxonomy[75] <- "p_2534_18B5"
heifer_test_fam2 <- merge(samples_heifer, taxa_heifer, by="row.names")
heifer_test_fam3 <- heifer_test_fam2[,-c(1,111,112,113,114)]
write.table(heifer_test_fam3, sep="\t", file="r_inputs/heifer_test_family_final.txt", row.names=F, col.names=T, quote=F)
#steer data
steer_test_family <- read.table("r_inputs/otu_table_test_steer_L5.txt", sep="\t", header=T)
samples_steer <- steer_test_family[,-1]
taxa_steer <- as.data.frame(str_split_fixed(steer_test_family$taxonomy, ";", 5))
names(taxa_steer)[5] <- "taxonomy"
taxa_steer$taxonomy <- sub("f__", "", taxa_steer$taxonomy)
taxa_steer$taxonomy <- sub("\\]", "", taxa_steer$taxonomy)
taxa_steer$taxonomy <- sub("\\[", "", taxa_steer$taxonomy)
taxa_steer$taxonomy <- sub("^$", "No Assigned Family", taxa_steer$taxonomy)
taxa_steer$taxonomy <- sub("Other", "No Assigned Family", taxa_steer$taxonomy)
taxa_steer$taxonomy[1] <- "NoAssignedFamily1"
taxa_steer$taxonomy[60] <- "NoAssignedFamily2"
taxa_steer$taxonomy[61] <- "NoAssignedFamily3"
taxa_steer$taxonomy[114] <- "NoAssignedFamily4"
taxa_steer$taxonomy[134] <- "NoAssignedFamily5"
taxa_steer$taxonomy[135] <- "NoAssignedFamily6"
taxa_steer$taxonomy[165] <- "NoAssignedFamily7"
taxa_steer$taxonomy[183] <- "NoAssignedFamily8"
taxa_steer$taxonomy[188] <- "NoAssignedFamily9"
taxa_steer$taxonomy[223] <- "NoAssignedFamily10"
taxa_steer$taxonomy[228] <- "NoAssignedFamily11"
taxa_steer$taxonomy[241] <- "NoAssignedFamily12"
taxa_steer$taxonomy[264] <- "NoAssignedFamily13"
taxa_steer$taxonomy[265] <- "NoAssignedFamily14"
taxa_steer$taxonomy[284] <- "NoAssignedFamily15"
taxa_steer$taxonomy[286] <- "NoAssignedFamily16"
taxa_steer$taxonomy[288] <- "NoAssignedFamily17"
taxa_steer$taxonomy[290] <- "NoAssignedFamily18"
taxa_steer$taxonomy[291] <- "NoAssignedFamily19"
taxa_steer$taxonomy[306] <- "NoAssignedFamily20"
taxa_steer$taxonomy[71] <- "S24_7"
taxa_steer$taxonomy[75] <- "p_2534_18B5"
#validation set
taxa_steer$taxonomy[167] <- "R4_45B"
steer_test_fam2 <- merge(samples_steer, taxa_steer, by="row.names")
steer_test_fam3 <- steer_test_fam2[,-c(1,108,109,110,111)]
write.table(steer_test_fam3, sep="\t", file="r_inputs/steer_test_family_final.txt", row.names=F, col.names=T, quote=F)
```
## Arcsine square root transformation of test sets at family level for heifer and steer cohorts
```{r}
#Load packages
library(dplyr)
#Differential families sets
differential_fam_heifer <- read.table("differential_otus/heifer_fam_differential.txt", sep="\t", header=F)
differential_fam_steer <- read.table("differential_otus/steer_fam_differential.txt", sep="\t", header=F)
#heifer data
heifer_family <- read.table("r_inputs/heifer_test_family_final.txt", sep="\t", header=T, fill=T, stringsAsFactors=F)
heifer_taxa <- subset(heifer_family, select = c(taxonomy))
heifer_data <- heifer_family[,-110]
#arcsine square root transformation
trans.arcsine <- function(x){asin(sign(x) * sqrt(abs(x)))}
heifer_data[, 1:109] <- as.data.frame(lapply(heifer_data[, 1:109], FUN = function(x) {sapply(x, FUN = trans.arcsine)}))
merged_heifer <- merge(heifer_taxa, heifer_data, by = "row.names")
merged_heifer <- merged_heifer[,-1]
heifer_filtered <- merged_heifer[which(merged_heifer$taxonomy %in% differential_fam_heifer$V1),]
heifer_trans <- as.data.frame(t(heifer_filtered), stringsAsFactors=F)
colnames(heifer_trans) = heifer_trans[1, ]
heifer_trans = heifer_trans[-1, ]
heifer_trans_final <- add_rownames(heifer_trans, "SampleID")
write.table(heifer_trans_final, file="stepwise_reg/heifer_fam_final.txt", sep="\t", col.names=T, row.names=F, quote=F)
#steer data
steer_family <- read.table("r_inputs/steer_test_family_final.txt", sep="\t", header=T, fill=T, stringsAsFactors=F)
steer_taxa <- subset(steer_family, select = c(taxonomy))
steer_data <- steer_family[,-107]
#arcsine square root transformation
trans.arcsine <- function(x){asin(sign(x) * sqrt(abs(x)))}
steer_data[, 1:106] <- as.data.frame(lapply(steer_data[, 1:106], FUN = function(x) {sapply(x, FUN = trans.arcsine)}))
merged_steer <- merge(steer_taxa, steer_data, by = "row.names")
merged_steer <- merged_steer[,-1]
steer_filtered <- merged_steer[which(merged_steer$taxonomy %in% differential_fam_steer$V1),]
steer_trans <- as.data.frame(t(steer_filtered), stringsAsFactors=F)
colnames(steer_trans) = steer_trans[1, ]
steer_trans = steer_trans[-1, ]
steer_trans_final <- add_rownames(steer_trans, "SampleID")
write.table(steer_trans_final, file="stepwise_reg/steer_fam_final.txt", sep="\t", col.names=T, row.names=F, quote=F)
#Heifer validation set
validation_heifer <- merged_steer[which(merged_steer$taxonomy %in% differential_fam_heifer$V1),]
validation_heifer_trans <- as.data.frame(t(validation_heifer), stringsAsFactors=F)
colnames(validation_heifer_trans) = validation_heifer_trans[1, ]
validation_heifer_trans = validation_heifer_trans[-1, ]
validation_heifer_final <- add_rownames(validation_heifer_trans, "SampleID")
write.table(validation_heifer_final, file = "stepwise_reg/validation_heifer_fam.txt", sep="\t", col.names=T, row.names=F, quote=F)
#Steer validation set
validation_steer <- merged_heifer[which(merged_heifer$taxonomy %in% differential_fam_steer$V1),]
validation_steer_trans <- as.data.frame(t(validation_steer), stringsAsFactors=F)
colnames(validation_steer_trans) = validation_steer_trans[1, ]
validation_steer_trans = validation_steer_trans[-1, ]
validation_steer_final <- add_rownames(validation_steer_trans, "SampleID")
write.table(validation_steer_final, file = "stepwise_reg/validation_steer_fam.txt", sep="\t", col.names=T, row.names=F, quote=F)
```
#Forward stepwise regression of taxa-based models at family level for average daily feed intake (ADFI), average daily gain (ADG), feed efficiency (ADG/ADFI) within heifer and steer cohort
```{r}
#Load packages
library(car)
#Heifer data
mapping_heifer <- read.table("r_inputs/mapping_test_heifer.txt", sep="\t", header=T)
row.names(mapping_heifer) <- mapping_heifer[,1]
#Heifer linear models for ADFI, ADG, and FE using breed composition as covariates
lm_ADFI_heifer <- lm(ADFI ~ perANS + perHHS + perARS + perSHS + perBMS + perBRS + perBNS + perSGS + perBVS + perCHS + perCAS + perGVS + perLMS + perMAS + perSAS + perSMS + perHH + perAN + perSM + perCH + perM2 + perM3 + perRS + perRO, data=mapping_heifer)
ADFI_res_heifer <- as.data.frame(resid(lm_ADFI_heifer))
lm_ADG_heifer <- lm(ADG ~ perANS + perHHS + perARS + perSHS + perBMS + perBRS + perBNS + perSGS + perBVS + perCHS + perCAS + perGVS + perLMS + perMAS + perSAS + perSMS + perHH + perAN + perSM + perCH + perM2 + perM3 + perRS + perRO, data=mapping_heifer)
ADG_res_heifer <- as.data.frame(resid(lm_ADG_heifer))
lm_FE_heifer <- lm(FE ~ perANS + perHHS + perARS + perSHS + perBMS + perBRS + perBNS + perSGS + perBVS + perCHS + perCAS + perGVS + perLMS + perMAS + perSAS + perSMS + perHH + perAN + perSM + perCH + perM2 + perM3 + perRS + perRO, data=mapping_heifer)
FE_res_heifer <- as.data.frame(resid(lm_FE_heifer))
#Heifer dataframes
heifer_fam <- read.table("stepwise_reg/heifer_fam_final.txt", sep = "\t", header=T)
row.names(heifer_fam) <- heifer_fam[,1]
heifer_fam <- heifer_fam[,-1]
heifer_ADFI <- merge(ADFI_res_heifer,heifer_fam,by="row.names")
heifer_ADFI <- heifer_ADFI[,-1]
colnames(heifer_ADFI)[1] <- "ADFI"
heifer_ADG <- merge(ADG_res_heifer,heifer_fam,by="row.names")
heifer_ADG <- heifer_ADG[,-1]
colnames(heifer_ADG)[1] <- "ADG"
heifer_FE <- merge(FE_res_heifer,heifer_fam,by="row.names")
heifer_FE <- heifer_FE[,-1]
colnames(heifer_FE)[1] <- "FE"
#Heifer models
#stepwise regression forward ADFI
null_heifer_ADFI_fam=lm(ADFI~1, data=heifer_ADFI)
#summary(null_heifer_ADFI_fam)
full_heifer_ADFI_fam=lm(ADFI~., data=heifer_ADFI)
#summary(full_heifer_ADFI_fam)
step(null_heifer_ADFI_fam, scope=list(lower=null_heifer_ADFI_fam, upper=full_heifer_ADFI_fam), direction="forward")
#Model ADFI
model_heifer_ADFI_fam <- lm(formula = ADFI ~ NoAssignedFamily18 + Veillonellaceae + Coriobacteriaceae, data=heifer_ADFI)
summary(model_heifer_ADFI_fam)
##anova(model_heifer_ADFI_fam)
#Test multi-collinearity (variance inflation factor)
vif(model_heifer_ADFI_fam)
sqrt(vif(model_heifer_ADFI_fam)) > 2
#Homoscedasticity & Normality
layout(matrix(c(1,2,3,4),2,2))
plot(model_heifer_ADFI_fam)
#Extract observed and predicted values
OP_heifer_ADFI_fam <- data.frame(Observed = heifer_ADFI$ADFI, Predicted = fitted(model_heifer_ADFI_fam))
#Plot observed vs prediceted
par(mfrow=c(1,1))
plot(OP_heifer_ADFI_fam$Predicted, OP_heifer_ADFI_fam$Observed, xlab="Predicted Average Daily Feed Intake (kg/d)", ylab="Observed Average Daily Feed Intake (kg/d)")
OP=lm(Observed~Predicted, data=OP_heifer_ADFI_fam)
#summary(OP)
abline(OP)
#stepwise regression forward ADG
null_heifer_ADG_fam=lm(ADG~1, data=heifer_ADG)
#summary(null_heifer_ADG_fam)
full_heifer_ADG_fam=lm(ADG~., data=heifer_ADG)
#summary(full_heifer_ADG_fam)
step(null_heifer_ADG_fam, scope=list(lower=null_heifer_ADG_fam, upper=full_heifer_ADG_fam), direction="forward")
#Model ADG
model_heifer_ADG_fam <- lm(formula = ADG ~ Veillonellaceae + Coriobacteriaceae, data=heifer_ADG)
summary(model_heifer_ADG_fam)
#anova(model_heifer_ADG_fam)
#Test multi-collinearity (variance inflation factor)
vif(model_heifer_ADG_fam)
sqrt(vif(model_heifer_ADG_fam)) > 2
#Homoscedasticity & Normality
layout(matrix(c(1,2,3,4),2,2))
plot(model_heifer_ADG_fam)
#Extract observed and predicted values
OP_heifer_ADG_fam <- data.frame(Observed = heifer_ADG$ADG, Predicted = fitted(model_heifer_ADG_fam))
#Plot observed vs prediceted
par(mfrow=c(1,1))
plot(OP_heifer_ADG_fam$Predicted, OP_heifer_ADG_fam$Observed, xlab="Predicted Average Daily Gain (kg/d)", ylab="Observed Average Daily Gain (kg/d)")
OP=lm(Observed~Predicted, data=OP_heifer_ADG_fam)
#summary(OP)
abline(OP)
#stepwise regression forward FE
null_heifer_FE_fam=lm(FE~1, data=heifer_FE)
#summary(null_heifer_FE_fam)
full_heifer_FE_fam=lm(FE~., data=heifer_FE)
#summary(full_heifer_FE_fam)
step(null_heifer_FE_fam, scope=list(lower=null_heifer_FE_fam, upper=full_heifer_FE_fam), direction="forward")
#Model FE
model_heifer_FE_fam <- lm(formula = FE ~ Veillonellaceae + Coriobacteriaceae + NoAssignedFamily2, data=heifer_FE)
summary(model_heifer_FE_fam)
#anova(model_heifer_FE_fam)
#Test multi-collinearity (variance inflation factor)
vif(model_heifer_FE_fam)
sqrt(vif(model_heifer_FE_fam)) > 2
#Homoscedasticity & Normality
layout(matrix(c(1,2,3,4),2,2))
plot(model_heifer_FE_fam)
#Extract observed and predicted values
OP_heifer_FE_fam <- data.frame(Observed = heifer_FE$FE, Predicted = fitted(model_heifer_FE_fam))
#Plot observed vs prediceted
par(mfrow=c(1,1))
plot(OP_heifer_FE_fam$Predicted, OP_heifer_FE_fam$Observed, xlab="Predicted Feed Efficiency (ADG/ADFI)", ylab="Observed Feed Efficiency (ADG/ADFI)")
OP=lm(Observed~Predicted, data=OP_heifer_FE_fam)
##summary(OP)
abline(OP)
#Steer data
mapping_steer <- read.table("r_inputs/mapping_test_steer.txt", sep="\t", header=T)
row.names(mapping_steer) <- mapping_steer[,1]
#Steer linear models for ADFI, ADG, and FE using breed composition as covariates
lm_ADFI_steer <- lm(ADFI ~ perANS + perHHS + perARS + perSHS + perBMS + perBRS + perBNS + perSGS + perBVS + perCHS + perCAS + perGVS + perLMS + perMAS + perSAS + perSMS + perHH + perAN + perSM + perCH + perM2 + perM3 + perRS + perRO, data=mapping_steer)
ADFI_res_steer <- as.data.frame(resid(lm_ADFI_steer))
lm_ADG_steer <- lm(ADG ~ perANS + perHHS + perARS + perSHS + perBMS + perBRS + perBNS + perSGS + perBVS + perCHS + perCAS + perGVS + perLMS + perMAS + perSAS + perSMS + perHH + perAN + perSM + perCH + perM2 + perM3 + perRS + perRO, data=mapping_steer)
ADG_res_steer <- as.data.frame(resid(lm_ADG_steer))
lm_FE_steer <- lm(FE ~ perANS + perHHS + perARS + perSHS + perBMS + perBRS + perBNS + perSGS + perBVS + perCHS + perCAS + perGVS + perLMS + perMAS + perSAS + perSMS + perHH + perAN + perSM + perCH + perM2 + perM3 + perRS + perRO, data=mapping_steer)
FE_res_steer <- as.data.frame(resid(lm_FE_steer))
#steer dataframes
steer_fam <- read.table("stepwise_reg/steer_fam_final.txt", sep="\t", header=T)
row.names(steer_fam) <- steer_fam[,1]
steer_fam <- steer_fam[,-1]
steer_ADFI <- merge(ADFI_res_steer,steer_fam,by="row.names")
steer_ADFI <- steer_ADFI[,-1]
colnames(steer_ADFI)[1] <- "ADFI"
steer_ADG <- merge(ADG_res_steer,steer_fam,by="row.names")
steer_ADG <- steer_ADG[,-1]
colnames(steer_ADG)[1] <- "ADG"
steer_FE <- merge(FE_res_steer,steer_fam,by="row.names")
steer_FE <- steer_FE[,-1]
colnames(steer_FE)[1] <- "FE"
#steer models
#stepwise regression forward ADFI
null_steer_ADFI_fam=lm(ADFI~1, data=steer_ADFI)
#summary(null_steer_ADFI_fam)
full_steer_ADFI_fam=lm(ADFI~., data=steer_ADFI)
#summary(full_steer_ADFI_fam)
step(null_steer_ADFI_fam, scope=list(lower=null_steer_ADFI_fam, upper=full_steer_ADFI_fam), direction="forward")
#Model ADFI
model_steer_ADFI_fam <- lm(formula = ADFI ~ NoAssignedFamily13 + Paraprevotellaceae + Lachnospiraceae + Coriobacteriaceae + Bifidobacteriaceae + Erysipelotrichaceae, data=steer_ADFI)
summary(model_steer_ADFI_fam)
##anova(model_steer_ADFI_fam)
#Test multi-collinearity (variance inflation factor)
vif(model_steer_ADFI_fam)
sqrt(vif(model_steer_ADFI_fam)) > 2
#Homoscedasticity & Normality
layout(matrix(c(1,2,3,4),2,2))
plot(model_steer_ADFI_fam)
#Extract observed and predicted values
OP_steer_ADFI_fam <- data.frame(Observed = steer_ADFI$ADFI, Predicted = fitted(model_steer_ADFI_fam))
#Plot observed vs prediceted
par(mfrow=c(1,1))
plot(OP_steer_ADFI_fam$Predicted, OP_steer_ADFI_fam$Observed, xlab="Predicted Average Daily Feed Intake (kg/d)", ylab="Observed Average Daily Feed Intake (kg/d)")
OP=lm(Observed~Predicted, data=OP_steer_ADFI_fam)
#summary(OP)
abline(OP)
#stepwise regression forward ADG
null_steer_ADG_fam=lm(ADG~1, data=steer_ADG)
#summary(null_steer_ADG_fam)
full_steer_ADG_fam=lm(ADG~., data=steer_ADG)
#summary(full_steer_ADG_fam)
step(null_steer_ADG_fam, scope=list(lower=null_steer_ADG_fam, upper=full_steer_ADG_fam), direction="forward")
#Model ADG
model_steer_ADG_fam <- lm(formula = ADG ~ p_2534_18B5 + Lachnospiraceae + Coriobacteriaceae + Bifidobacteriaceae + NoAssignedFamily9, data=steer_ADG)
summary(model_steer_ADG_fam)
#anova(model_steer_ADG_fam)
#Test multi-collinearity (variance inflation factor)
vif(model_steer_ADG_fam)
sqrt(vif(model_steer_ADG_fam)) > 2
#Homoscedasticity & Normality
layout(matrix(c(1,2,3,4),2,2))
plot(model_steer_ADG_fam)
#Extract observed and predicted values
OP_steer_ADG_fam <- data.frame(Observed = steer_ADG$ADG, Predicted = fitted(model_steer_ADG_fam))
#Plot observed vs prediceted
par(mfrow=c(1,1))
plot(OP_steer_ADG_fam$Predicted, OP_steer_ADG_fam$Observed, xlab="Predicted Average Daily Gain (kg/d)", ylab="Observed Average Daily Gain (kg/d)")
OP=lm(Observed~Predicted, data=OP_steer_ADG_fam)
#summary(OP)
abline(OP)
#stepwise regression forward FE
null_steer_FE_fam=lm(FE~1, data=steer_FE)
#summary(null_steer_FE_fam)
full_steer_FE_fam=lm(FE~., data=steer_FE)
#summary(full_steer_FE_fam)
step(null_steer_FE_fam, scope=list(lower=null_steer_FE_fam, upper=full_steer_FE_fam), direction="forward")
#Model FE
model_steer_FE_fam <- lm(formula = FE ~ NoAssignedFamily13 + Paraprevotellaceae + NoAssignedFamily9, data=steer_FE)
summary(model_steer_FE_fam)
#anova(model_steer_FE_fam)
#Test multi-collinearity (variance inflation factor)
vif(model_steer_FE_fam)
sqrt(vif(model_steer_FE_fam)) > 2
#Homoscedasticity & Normality
layout(matrix(c(1,2,3,4),2,2))
plot(model_steer_FE_fam)
#Extract observed and predicted values
OP_steer_FE_fam <- data.frame(Observed = steer_FE$FE, Predicted = fitted(model_steer_FE_fam))
#Plot observed vs prediceted
par(mfrow=c(1,1))
plot(OP_steer_FE_fam$Predicted, OP_steer_FE_fam$Observed, xlab="Predicted Feed Efficiency (ADG/ADFI)", ylab="Observed Feed Efficiency (ADG/ADFI)")
OP=lm(Observed~Predicted, data=OP_steer_FE_fam)
#summary(OP)
abline(OP)
#Validation heifer data sets
validation_fam_heifer <- read.table("stepwise_reg/validation_heifer_fam.txt", sep="\t", header=T)
row.names(validation_fam_heifer) <- validation_fam_heifer[,1]
validation_fam_heifer <- validation_fam_heifer[,-1]
validation_heifer_fam_ADFI <- merge(ADFI_res_steer,validation_fam_heifer,by="row.names")
validation_heifer_fam_ADFI <- validation_heifer_fam_ADFI[,-1]
colnames(validation_heifer_fam_ADFI)[1] <- "ADFI"
validation_heifer_fam_ADG <- merge(ADG_res_steer,validation_fam_heifer,by="row.names")
validation_heifer_fam_ADG <- validation_heifer_fam_ADG[,-1]
colnames(validation_heifer_fam_ADG)[1] <- "ADG"
validation_heifer_fam_FE <- merge(FE_res_steer,validation_fam_heifer,by="row.names")
validation_heifer_fam_FE <- validation_heifer_fam_FE[,-1]
colnames(validation_heifer_fam_FE)[1] <- "FE"
#Validation heifer models
#Heifer ADFI
modelval_heifer_fam_ADFI <- lm(formula = ADFI ~ NoAssignedFamily18 + Veillonellaceae + Coriobacteriaceae, data=validation_heifer_fam_ADFI)
summary(modelval_heifer_fam_ADFI)
#Heifer ADG
modelval_heifer_fam_ADG <- lm(formula = ADG ~ Veillonellaceae + Coriobacteriaceae, data=validation_heifer_fam_ADG)
summary(modelval_heifer_fam_ADG)
#Heifer FE
modelval_heifer_fam_FE <- lm(formula = FE ~ Veillonellaceae + Coriobacteriaceae + NoAssignedFamily2, data=validation_heifer_fam_FE)
summary(modelval_heifer_fam_FE)
#Validation steer data sets
validation_fam_steer <- read.table("stepwise_reg/validation_steer_fam.txt", sep="\t", header=T)
row.names(validation_fam_steer) <- validation_fam_steer[,1]
validation_fam_steer <- validation_fam_steer[,-1]
validation_steer_fam_ADFI <- merge(ADFI_res_heifer,validation_fam_steer,by="row.names")
validation_steer_fam_ADFI <- validation_steer_fam_ADFI[,-1]
colnames(validation_steer_fam_ADFI)[1] <- "ADFI"
validation_steer_fam_ADG <- merge(ADG_res_heifer,validation_fam_steer,by="row.names")
validation_steer_fam_ADG <- validation_steer_fam_ADG[,-1]
colnames(validation_steer_fam_ADG)[1] <- "ADG"
validation_steer_fam_FE <- merge(FE_res_heifer,validation_fam_steer,by="row.names")
validation_steer_fam_FE <- validation_steer_fam_FE[,-1]
colnames(validation_steer_fam_FE)[1] <- "FE"
#Validation steer models
#steer ADFI
modelval_steer_fam_ADFI <- lm(formula = ADFI ~ NoAssignedFamily13 + Paraprevotellaceae + Lachnospiraceae + Coriobacteriaceae + Bifidobacteriaceae + Erysipelotrichaceae, data=validation_steer_fam_ADFI)
summary(modelval_steer_fam_ADFI)
#steer ADG
modelval_steer_fam_ADG <- lm(formula = ADG ~ p_2534_18B5 + Lachnospiraceae + Coriobacteriaceae + Bifidobacteriaceae + NoAssignedFamily9, data=validation_steer_fam_ADG)
summary(modelval_steer_fam_ADG)
#steer FE
modelval_steer_fam_FE <- lm(formula = FE ~ NoAssignedFamily13 + Paraprevotellaceae + NoAssignedFamily9, data=validation_steer_fam_FE)
summary(modelval_steer_fam_FE)
```