-
Notifications
You must be signed in to change notification settings - Fork 4
/
plotAQuA_Contactmaps_Virtual4C.R
178 lines (136 loc) · 13.4 KB
/
plotAQuA_Contactmaps_Virtual4C.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
###HiC analysis, setup for compatibility with HiC-pro and Juicebox/juicer
### 1.Set up project and samples. Configure variables to adopt code into your own file system.
setwd("K:/projects/HiC/projects/")
project.title = "RH4_H3K27ac_HiChIP"
project.folder = paste(project.title,"/HiCpro_OUTPUT/hic_results/data/",sep="")
sample.list = list.dirs(path = project.folder, full.names = F, recursive = F)
project.data.folder = paste(project.title,"/HiCpro_OUTPUT/hic_results/data/",sep="")
project.title.mm10 = paste(project.title,"_mm10",sep="")
project.folder.mm10 = paste(project.title.mm10,"/HiCpro_OUTPUT/hic_results/data/",sep="") #folder changed to "stats" in newer HiC Pro pipeline for mergestat
sample.list.mm10 = list.dirs(path = project.folder.mm10, full.names = F, recursive = F)
project.data.folder.mm10 = paste(project.title.mm10,"/HiCpro_OUTPUT/hic_results/data/",sep="")
### 2. Obtain spike in read counts
mergestat.all = as.data.frame(read.table(paste(project.folder,"/",sample.list[1],"/",sample.list[1],"_allValidPairs.mergestat",sep=""), sep="\t", header=F)); mergestat.all = as.data.frame(mergestat.all$V1)
lapply(sample.list, function(x) {
##load and merge sample data
mergestat <- read.table(paste(project.folder,"/",x,"/",x,"_allValidPairs.mergestat",sep=""), sep="\t", header=F)
mergestat.sample = as.data.frame(mergestat[,2])
removable.string = "Sample_" ; sample.name = gsub(removable.string,"",x) ; colnames(mergestat.sample) = c(sample.name)
mergestat.all <<- cbind(mergestat.all,mergestat.sample)
})
lapply(sample.list.mm10, function(x) {
##load and merge sample data
mergestat <- read.table(paste(project.folder.mm10,"/",x,"/",x,"_allValidPairs.mergestat",sep=""), sep="\t", header=F)
mergestat.sample = as.data.frame(mergestat[,2])
removable.string = "Sample_" ; sample.name = gsub(removable.string,"mm10_",x) ; colnames(mergestat.sample) = c(sample.name)
mergestat.all <<- cbind(mergestat.all,mergestat.sample)
})
lapply(sample.list, function(x) {
##get percentage hg19 and mm10
removable.string = "Sample_" ; hg19 = gsub(removable.string,"",x) ; mm10 = gsub(removable.string,"mm10_",x)
mergestat.prct = as.data.frame(mergestat.all[,c(hg19,mm10)])
mergestat.prct$prct_hg19 = mergestat.prct[,1]/(mergestat.prct[,1]+mergestat.prct[,2])
mergestat.prct$prct_mm10 = mergestat.prct[,2]/(mergestat.prct[,1]+mergestat.prct[,2])
colnames(mergestat.prct) = paste(colnames(mergestat.prct),x,sep="_")
mergestat.all <<- cbind(mergestat.all,mergestat.prct[,3:4])
})
write.table(mergestat.all, paste("stats/",project.title,".mergestat.txt", sep=""),sep="\t", col.names = T, row.names = F)
### 3. Load and transform sparse matrix (from juicer: java -jar /usr/local/apps/juicer/juicer-1.5.6/scripts/juicer_tools.jar dump)
#if HiCcompare isn't installed, remove comments from the next 2 lines of code and run them.
#source("https://bioconductor.org/biocLite.R")
#biocLite("HiCcompare")
library(HiCcompare)
sample1 = "Sample_RH4_D6_H3K27ac_HiChIP_HKJ22BGX7";sample2 = "Sample_RH4_Ent6_H3K27ac_HiChIP_HKJ22BGX7"
removable.string = "Sample_" ; name1 = gsub(removable.string,"",sample1) ; name2 = gsub(removable.string,"",sample2)
mm10_name1 = gsub(removable.string,"mm10_",sample1) ; mm10_name2 = gsub(removable.string,"mm10_",sample2)
matrix.suffix = ".chr11.176MB-178MB.5KB.matrix.txt"
sparMat1 = read.table(paste(project.data.folder,sample1,"/",sample1,matrix.suffix,sep=""), header=F, sep="\t"); sparMat2 = read.table(paste(project.data.folder,sample2,"/",sample2,matrix.suffix,sep=""), header=F, sep="\t")
denMat1 <- sparse2full(sparMat1, hic.table = FALSE, column.name = NA); denMat2 <- sparse2full(sparMat2, hic.table = FALSE, column.name = NA)
denMat1.CPM = denMat1*1000000/(mergestat.all[2,name1]+mergestat.all[2,mm10_name1]) ; denMat2.CPM = denMat2*1000000/(mergestat.all[2,name2]+mergestat.all[2,mm10_name2])
AQuAfactor1 = mergestat.all[2,name1]/mergestat.all[2,mm10_name1] ; AQuAfactor2 = mergestat.all[2,name2]/mergestat.all[2,mm10_name2]
denAQuA1 = denMat1.CPM*AQuAfactor1; denAQuA2 = denMat2.CPM*AQuAfactor2
### 4. Build matrices with AQuA maximums, plot HEATMAPS
library(pheatmap)
library(ggplot2)
library(gridExtra)
quant_cut = 0.98 #caps the contact map plot values at a given percentile
AQuAmax = max(quantile(denAQuA1, probs = c(quant_cut)),quantile(denAQuA2, probs = c(quant_cut)))
denAQuA1max = denAQuA1; denAQuA1max[denAQuA1max>AQuAmax] <- AQuAmax
denAQuA2max = denAQuA2; denAQuA2max[denAQuA2max>AQuAmax] <- AQuAmax
denDeltaAQuA = denAQuA2 - denAQuA1;
DeltaAQuA.max = AQuAmax*0.5
denDeltaAQuA.max = denDeltaAQuA; denDeltaAQuA.max[denDeltaAQuA.max>DeltaAQuA.max] <- DeltaAQuA.max
denDeltaAQuA.max[denDeltaAQuA.max<(-DeltaAQuA.max)] <- (-DeltaAQuA.max)
breakList = seq(-(DeltaAQuA.max*1.05), (DeltaAQuA.max*1.05), by = AQuAmax/100)
#combine the plots!
p1 = pheatmap((denAQuA1max), cluster_rows = F, cluster_cols = F, color = colorRampPalette(c("white", "red"))(50))
p2 = pheatmap((denAQuA2max), cluster_rows = F, cluster_cols = F, color = colorRampPalette(c("white", "red"))(50))
p3 = pheatmap(denDeltaAQuA.max , cluster_rows = F, cluster_cols = F, breaks = breakList, color = colorRampPalette(c("dodgerblue", "white", "mediumvioletred"))(length(breakList)))
grid.arrange(arrangeGrob(p1[[4]],p2[[4]],p3[[4]],nrow=1,top = paste("AQuA: ",name1,", ",name2,", and delta from *",matrix.suffix,sep="")))
#save it!!
g <- arrangeGrob(p1[[4]],p2[[4]],p3[[4]],nrow=1,top = paste("AQuA: ",name1,", ",name2,", and delta from *",matrix.suffix,sep=""))
ggsave(file=paste(project.data.folder,sample2,"/",sample2,matrix.suffix,".3AQuAplots.pdf",sep=""), g, width = 14, height = 4.5) #saves g
#repeat with no AQuA
CPMmax = max(quantile(denMat1.CPM, probs = c(quant_cut)),quantile(denMat2.CPM, probs = c(quant_cut)))
denMat1.CPMmax = denMat1.CPM; denMat1.CPMmax[denMat1.CPMmax>CPMmax] <- CPMmax
denMat2.CPMmax = denMat2.CPM; denMat2.CPMmax[denMat2.CPMmax>CPMmax] <- CPMmax
denDelta.CPM = denMat2.CPM - denMat1.CPM
Delta.CPM.max = CPMmax*0.8
denDelta.CPM.max = denDelta.CPM; denDelta.CPM.max[denDelta.CPM.max>Delta.CPM.max] <- Delta.CPM.max
denDelta.CPM.max[denDelta.CPM.max<(-Delta.CPM.max)] <- (-Delta.CPM.max)
breakList.CPM = seq(-(Delta.CPM.max*1.05), (Delta.CPM.max*1.05), by = CPMmax/100)
#combine the plots!
p1 = pheatmap((denMat1.CPMmax), cluster_rows = F, cluster_cols = F, color = colorRampPalette(c("white", "red"))(50))
p2 = pheatmap((denMat2.CPMmax), cluster_rows = F, cluster_cols = F, color = colorRampPalette(c("white", "red"))(50))
p3 = pheatmap(denDelta.CPM.max, cluster_rows = F, cluster_cols = F, breaks = breakList.CPM, color = colorRampPalette(c("dodgerblue", "white", "mediumvioletred"))(length(breakList.CPM)))
grid.arrange(arrangeGrob(p1[[4]],p2[[4]],p3[[4]],nrow=1,top = paste("HiChIP: ",name1,", ",name2,", and delta from *",matrix.suffix,sep="")))
#save it!!
g <- arrangeGrob(p1[[4]],p2[[4]],p3[[4]],nrow=1,top = paste("HiChIP: ",name1,", ",name2,", and delta from *",matrix.suffix,sep=""))
ggsave(file=paste(project.data.folder,sample2,"/",sample2,matrix.suffix,".3plots.CPM.pdf",sep=""), g, width = 14, height = 4.5) #saves g
### 5. Extract a virtual 4C viewpoint, make bedgraphs, plot VIRTUAL4C
virt4C_viewpoint_chr = 'chr16'
virt4C_viewpoint = '960000'
df_AQuA1 = as.data.frame(denAQuA1, row.names = row.names(denAQuA1), col.names = col.names(denAQuA1))
df_AQuA2 = as.data.frame(denAQuA2, row.names = row.names(denAQuA2), col.names = col.names(denAQuA2))
virt4C.bedgraph1 = as.data.frame(df_AQuA1[,virt4C_viewpoint])
colnames(virt4C.bedgraph1) = "AQuA_contact_freq"
virt4C.bedgraph1$chr = virt4C_viewpoint_chr
virt4C.bedgraph1$start = as.numeric(rownames(df_AQuA1))
virt4C_binsize = virt4C.bedgraph1[2,c('start')] - virt4C.bedgraph1[1,c('start')]
virt4C.bedgraph1$stop = virt4C.bedgraph1$start + virt4C_binsize
virt4C.bedgraph1 = virt4C.bedgraph1[,c(2,3,4,1)]
virt4C.bedgraph2 = as.data.frame(df_AQuA2[,virt4C_viewpoint])
colnames(virt4C.bedgraph2) = "AQuA_contact_freq"
virt4C.bedgraph2$chr = virt4C_viewpoint_chr
virt4C.bedgraph2$start = as.numeric(rownames(df_AQuA2))
virt4C_binsize = virt4C.bedgraph2[2,c('start')] - virt4C.bedgraph2[1,c('start')]
virt4C.bedgraph2$stop = virt4C.bedgraph2$start + virt4C_binsize
virt4C.bedgraph2 = virt4C.bedgraph2[,c(2,3,4,1)]
window_coordinates = paste(virt4C_viewpoint_chr,":",virt4C.bedgraph1[1,c('start')],"-",max(virt4C.bedgraph1$stop),sep="")
## spline to smooth
spline1 <- as.data.frame(spline(virt4C.bedgraph1$start, virt4C.bedgraph1$AQuA_contact_freq))
spline2 <- as.data.frame(spline(virt4C.bedgraph2$start, virt4C.bedgraph2$AQuA_contact_freq))
spline_delta = spline1; spline_delta$y = (spline2$y-spline1$y)
spline_fold = spline1; spline_fold$y = (spline2$y+0.1)/(spline1$y+0.1) ; spline_fold = spline_fold[4:(nrow(spline_fold)-3),]
spline1.bedgraph = spline1; spline1.bedgraph[,1]=round(x = spline1.bedgraph[,1], digits = 0)+virt4C_binsize/2; spline1.bedgraph$chr = virt4C_viewpoint_chr ; spline1.bedgraph = spline1.bedgraph[,c(3,1,1,2)]
spline2.bedgraph = spline2; spline2.bedgraph[,1]=round(x = spline2.bedgraph[,1], digits = 0)+virt4C_binsize/2; spline2.bedgraph$chr = virt4C_viewpoint_chr ; spline2.bedgraph = spline2.bedgraph[,c(3,1,1,2)]
spline_delta.bedgraph = spline_delta; spline_delta.bedgraph[,1]=round(x = spline_delta.bedgraph[,1], digits = 0)+virt4C_binsize/2; spline_delta.bedgraph$chr = virt4C_viewpoint_chr ; spline_delta.bedgraph = spline_delta.bedgraph[,c(3,1,1,2)]
spline_fold.bedgraph = spline_fold; spline_fold.bedgraph[,1]=round(x = spline_fold.bedgraph[,1], digits = 0)+virt4C_binsize/2; spline_fold.bedgraph$chr = virt4C_viewpoint_chr ; spline_fold.bedgraph = spline_fold.bedgraph[,c(3,1,1,2)]
## write out bedgraphs
write.table(virt4C.bedgraph1, file=paste(project.data.folder,sample1,virt4C_viewpoint_chr,virt4C_viewpoint,".bedgraph",sep=""), sep="\t", row.names=FALSE, col.names=F, quote=FALSE)
write.table(virt4C.bedgraph2, file=paste(project.data.folder,sample2,virt4C_viewpoint_chr,virt4C_viewpoint,".bedgraph",sep=""), sep="\t", row.names=FALSE, col.names=F, quote=FALSE)
write.table(spline1.bedgraph, file=paste(project.data.folder,sample1,virt4C_viewpoint_chr,virt4C_viewpoint,".spline.bedgraph",sep=""), sep="\t", row.names=FALSE, col.names=F, quote=FALSE)
write.table(spline2.bedgraph, file=paste(project.data.folder,sample2,virt4C_viewpoint_chr,virt4C_viewpoint,".spline.bedgraph",sep=""), sep="\t", row.names=FALSE, col.names=F, quote=FALSE)
write.table(spline_delta.bedgraph, file=paste(project.data.folder,sample1,"_",sample2,virt4C_viewpoint_chr,virt4C_viewpoint,".spline_delta.bedgraph",sep=""), sep="\t", row.names=FALSE, col.names=F, quote=FALSE)
write.table(spline_fold.bedgraph, file=paste(project.data.folder,sample1,"_",sample2,virt4C_viewpoint_chr,virt4C_viewpoint,".spline_fold.bedgraph",sep=""), sep="\t", row.names=FALSE, col.names=F, quote=FALSE)
## make some plots
library(ggplot2)
ggplot(virt4C.bedgraph1, aes(x=(start+virt4C_binsize/2),y=AQuA_contact_freq))+geom_line()+geom_line(data=virt4C.bedgraph2, color = 'red')+
theme_bw()+geom_vline(xintercept = as.numeric(virt4C_viewpoint),color = "blue")+geom_vline(xintercept = (as.numeric(virt4C_viewpoint)+virt4C_binsize),color = "blue")
ggplot(spline1, aes(x=(x+virt4C_binsize/2),y=y))+geom_line()+geom_line(data=spline2, color = 'red')+
theme_bw()+geom_vline(xintercept = as.numeric(virt4C_viewpoint),color = "blue")+geom_vline(xintercept = (as.numeric(virt4C_viewpoint)+virt4C_binsize),color = "blue")
ggplot(spline1, aes(x=(x+virt4C_binsize/2),y=y))+geom_line(data=spline1, color = 'red')+geom_line(data=spline_delta, color = 'purple')+
theme_bw()+geom_vline(xintercept = as.numeric(virt4C_viewpoint),color = "blue")+geom_vline(xintercept = (as.numeric(virt4C_viewpoint)+virt4C_binsize),color = "blue")
#check spline coordinates by overlapping with raw bedgraph
ggplot(virt4C.bedgraph1, aes(x=(start+virt4C_binsize/2),y=AQuA_contact_freq))+geom_line()+geom_line(data=spline1, aes(x=x,y=y), color = 'red')+
theme_bw()+geom_vline(xintercept = as.numeric(virt4C_viewpoint),color = "blue")+geom_vline(xintercept = (as.numeric(virt4C_viewpoint)+virt4C_binsize),color = "blue")