-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathCIA.R
191 lines (154 loc) · 6.3 KB
/
CIA.R
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
setwd(dirname(rstudioapi::getActiveDocumentContext()$path)) #setwd to directory where script is located
#index of divergent and balancing SNPs
index.div.alc <- c(98,99,109, #bayescan
36, #lfmm altitude (98 was double)
84,81,79,126,67,41,47, 26,23,14, 65,28,40,51,69)
index.bal.alc <- c(29,25,116,108,7)
index.div.gen <- c(5,23,29,59, #bayescan
39,82) #lfmm (23 was double)
index.bal.gen <- c(21,53,84,72,77,73,80,87,88,32,97)
#if (!requireNamespace("BiocManager", quietly = TRUE)) install.packages("BiocManager")
#BiocManager::install("made4")
library(made4)
AF.gen <- read.table('gentianAF', h = T , sep= " ")
AF.alc <- read.table('alconAF', h = T, sep = " ")
x <- rownames(AF.alc[rownames(AF.alc)%in%rownames(AF.gen),])
#rownames(AF.gen[rownames(AF.gen)%in%rownames(AF.alc),])
#data of 24 populations is available for both species
#select only these and transpose
?cia #genes as rows, pops as column
AF.gen <- data.frame(t(AF.gen[x,]))
AF.alc <- data.frame(t(AF.alc[x,]))
cia(AF.gen, AF.alc) #RV 0.74 (total correlation quite strong as expected)
#balancing only
cia.ABal.GBal <- cia(AF.gen[index.bal.gen,], AF.alc[index.bal.alc,])
cia.ABal.GBal$coinertia$RV #0.270
#Balancing - Neutral
cia.Abal.Gneu <- cia(AF.alc[index.bal.alc,],
AF.gen[-(c(index.bal.gen,index.div.gen)),])
cia.Abal.Gneu$coinertia$RV #0.256
cia.Aneu.Gbal <- cia(AF.alc[-(c(index.bal.alc,index.div.alc)),],
AF.gen[index.bal.gen,])
cia.Aneu.Gbal$coinertia$RV #0.419 (highest correlation)
#selecting retained axes based on scree plot (cia.scan = T) gives same results
#divergent only
cia.Adiv.Gdiv <- cia(AF.gen[index.div.gen,], AF.alc[index.div.alc,])
cia.Adiv.Gdiv$coinertia$RV #0.506
#divergent - neutral
cia.Adiv.Gneu <- cia(AF.alc[index.div.alc,],
AF.gen[-(c(index.bal.gen,index.div.gen)),])
cia.Adiv.Gneu$coinertia$RV #0.636
cia.Aneu.Gdiv <- cia(AF.alc[-(c(index.bal.alc,index.div.alc)),],
AF.gen[index.div.gen,])
cia.Aneu.Gdiv$coinertia$RV #0.562
#divergent - balancing
cia.Adiv.Gbal <- cia(AF.alc[index.div.alc,], AF.gen[index.bal.gen,])
cia.Adiv.Gbal$coinertia$RV
cia.Abal.Gdiv <- cia(AF.alc[index.bal.alc,], AF.gen[index.div.gen,])
cia.Abal.Gdiv$coinertia$RV
#neutral - neutral
cia.Aneu.Gneu <- cia(AF.alc[-(c(index.bal.alc,index.div.alc)),],
AF.gen[-(c(index.bal.gen,index.div.gen)),])
cia.Aneu.Gneu$coinertia$RV
##Sample Sizes-> equal(11)
#load bayescan results
bayes.alc <- read.delim(file = '/home/jonas/Documents/Masterthesis/data/Analyses/Bayescan/BayescanOutput/alcon/alcon_fst.txt', h = T,sep = ' ')
bayes.alc <- bayes.alc[,3:7]
colnames(bayes.alc) <- c('prob','log10(PO)','qval','alpha','fst')
bayes.gen <- read.delim(file = '/home/jonas/Documents/Masterthesis/data/Analyses/Bayescan/BayescanOutput/gentian/gentian_fst.txt', h = T,sep = ' ')
bayes.gen <- bayes.gen[,3:7]
colnames(bayes.gen) <- c('prob','log10(PO)','qval','alpha','fst')
#select 11 most 'balanced' genes (lowest alpha)
rownames(bayes.alc[order(bayes.alc$alpha),])[1:11]
index.bal.alc
iba2 <- order(bayes.alc$alpha)[1:11]
ina2 <- setdiff(c(1:127), c(iba2,index.div.alc)) #too sample random neutral
ina3 <- order(abs(bayes.alc$alpha))[1:11] #11 snps with alpha closest to 0
ing2 <- setdiff(c(1:105),(c(index.bal.gen,index.div.gen)))
ing3 <- order(abs(bayes.gen$alpha))[1:11]
#11 balancing inertia
cia.ABal.GBal2 <- cia(AF.gen[index.bal.gen,], AF.alc[iba2,])
cia.ABal.GBal2$coinertia$RV #0.240
#random sample of 11 neutral
#balancing(alcon) - neutral(gentian)
cia.Abal.Gneu2 <- cia(AF.alc[iba2,],
AF.gen[sample(ing2, size=11),])
cia.Abal.Gneu2$coinertia$RV #0.28
res <- c()
for (i in 1:1000){
cia <- cia(AF.alc[iba2,],
AF.gen[sample(ing2, size=11),])
res[i] <- cia$coinertia$RV
}
c(mean(res), sd(res)) #0.297 +/- 0.054
#neutral(alcon) - balancing(gentian)
cia.Aneu.Gbal2 <- cia(AF.alc[sample(ina2, size =11),],
AF.gen[index.bal.gen,])
cia.Aneu.Gbal2$coinertia$RV #0.33 (highest correlation)
res1 <- c()
for (i in 1:1000){
cia <- cia(AF.alc[sample(ina2, size =11),],
AF.gen[index.bal.gen,])
res1[i] <- cia$coinertia$RV
}
c(mean(res1), sd(res1)) #0.309 +/- 0.052
#11 'most neutral' snps
cia.Abal.Gneu3 <- cia(AF.alc[iba2,],
AF.gen[ing3,])
cia.Abal.Gneu3$coinertia$RV #0.360
cia.Aneu.Gbal3 <- cia(AF.alc[ina3,],
AF.gen[index.bal.gen,])
cia.Aneu.Gbal3$coinertia$RV #0.355 (highest correlation)
?cia
####################
#instead of 6 extra alcon, only take 5 balanced gentian genes
ibg4 <- order(bayes.gen$alpha)[1:5]
cia.ABal.GBal4 <- cia(AF.alc[index.bal.alc,], AF.gen[ibg4,])
cia.ABal.GBal4$coinertia$RV #0.152
res3 <- c()
for (i in 1:500){
cia <- cia(AF.alc[sample(ina2, size =5),],
AF.gen[ibg4,])
res3[i] <- cia$coinertia$RV
}
c(mean(res3), sd(res3)) #0.123 +/- 0.046
res4 <- c()
for (i in 1:500){
cia <- cia(AF.alc[index.bal.alc,],
AF.gen[sample(ing2, size = 5),])
res4[i] <- cia$coinertia$RV
}
c(mean(res4), sd(res4)) #0.151 +/- 0.046
#similar results
####################
#divergent snps
##
idg <- order(bayes.gen$alpha)[(105-18):105]
#div - div
cia.Adiv.Gdiv <- cia(AF.alc[index.div.alc,], AF.gen[idg,])
cia.Adiv.Gdiv$coinertia$RV ##0.484
#Aneu-Gdiv
res5 <- c()
for (i in 1:500){
cia <- cia(AF.alc[sample(ina2, size =19),],
AF.gen[idg,])
res5[i] <- cia$coinertia$RV
}
c(mean(res5), sd(res5)) #0.520 +/- 0.061
#Adiv-Gneu
res6 <- c()
for (i in 1:500){
cia <- cia(AF.alc[index.div.alc,],
AF.gen[sample(ing2, size = 19),])
res6[i] <- cia$coinertia$RV
}
c(mean(res6), sd(res6)) #0.545 +/- 0.038
##via dudi.pca and coinertia() -> not working because of unequal rows
dudi.div.alc <- dudi.pca(AF.alc[index.div.alc,], scannf = FALSE, nf = 2)
dudi.bal.alc <- dudi.pca(AF.alc[index.bal.alc,], scannf = FALSE, nf = 2)
dudi.neu.alc <- dudi.pca(AF.alc[-(c(index.bal.alc,index.div.alc)),],scannf = FALSE, nf = 1)
dudi.div.gen <- dudi.pca(AF.gen[index.div.gen,], scannf = FALSE, nf = 3) #3 AF.genes was best
dudi.bal.gen <- dudi.pca(AF.gen[index.bal.gen,], scannf = FALSE, nf = 1) #also 3, although less clear
dudi.neu.gen <- dudi.pca(AF.gen[-(c(index.bal.gen,index.div.gen)),],
scannf = FALSE, nf = 1) #only 2 seems best
coinertia(dudi.bal.alc, dudi.bal.gen)