-
Notifications
You must be signed in to change notification settings - Fork 1
/
Copy pathsupplementary3_human_tbp.Rmd
181 lines (130 loc) · 6.49 KB
/
supplementary3_human_tbp.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
``` {r setup, echo=FALSE, message=FALSE, include=FALSE, error=FALSE}
library(GenomicRanges, warn.conflicts=F)
library(magrittr)
library(parallel)
library(ggplot2)
library(reshape)
setwd("/data/analysis_code/")
options(knitr.figure_dir = "supplementary3_human_tbp")
source("shared_code/knitr_common.r")
source("shared_code/granges_common.r")
source("shared_code/metapeak_common.r")
source("shared_code/sample_common.r")
source("shared_code/heatmap_common.r")
```
# Supplementary human TBP
**Author:** [Wanqing Shao](mailto:[email protected])
**Generated:** `r format(Sys.time(), "%a %b %d %Y, %I:%M %p")`
Plot TBP data in human cells
### Generating isolated TSS and correct TSS position based on K562 grocap data
Take Homo_sapiens.GRCh37.75.gtf.gz and use Lis's group's K562 grocap data to reannotate the TSS, then filter out start sites that are less than 300bp from each other
```{r eval=F}
library(BSgenome.Hsapiens.UCSC.hg19)
all.tx <- import("/data/public_data/human_annotation/Homo_sapiens.GRCh37.75.gtf.gz")
tx.gr <- all.tx[all.tx$type == "transcript" & all.tx$source == "protein_coding"]
seqlevels(tx.gr) <- paste0("chr", seqlevels(tx.gr))
cap.pos <- import("/data/public_data/lis_grocap/GSM1480321_K562_GROcap_wTAP_plus.bigWig", asRle = T)
cap.neg <- abs(import("/data/public_data/lis_grocap/GSM1480321_K562_GROcap_wTAP_minus.bigWig", asRle = T))
cap.pos <- cap.pos/ sum(cap.pos) * 1000000
cap.neg <- cap.neg/ sum(cap.neg) * 1000000
tx.gr <- tx.gr[seqnames(tx.gr) %in% seqlevels(cap.pos)]
tss_realignment <- function(strand, procap){
tx <- tx.gr[strand(tx.gr) == strand]
tx.r <- resize(tx, 1, "start") %>% resize(., 301, "center")
tx.r$new_start <- procap %>% regionWhichMaxs(tx.r,.)
tx.r$procap_sig <- procap %>% regionSums(tx.r, .)
start(tx.r) <- ifelse(tx.r$procap_sig >= 20, tx.r$new_start, start(tx.r))
end(tx.r) <- start(tx.r)
tx.r
}
tss_pos <- tss_realignment("+", cap.pos)
tss_neg <- tss_realignment("-", cap.neg)
tss <- c(tss_pos, tss_neg)
seqlevels(tss, force=T) <- c("chr1", "chr2", "chr3", "chr4", "chr5", "chr6", "chr7", "chr8",
"chr9", "chr10", "chr11", "chr12", "chr13", "chr14", "chr15",
"chr16", "chr17", "chr18", "chr19", "chr20", "chr21", "chr22",
"chrX", "chrY", "chrM")
tss_temp <- tss[order(tss$procap_sig, decreasing=T)] %>%
.[!duplicated(paste(seqnames(.), start(.)))]
strand(tss_temp) <- "*"
tss_temp <- tss_temp[order(tss_temp)]
all_dis <- c(start(tss_temp), Inf) - c(0, start(tss_temp))
tss_temp$dis_before <- all_dis[1:length(tss_temp)]
tss_temp$dis_after <- all_dis[2:length(all_dis)]
tss_temp <-subset(tss_temp, dis_before > 300 & dis_after > 300)
tss_u <- tss[tss$transcript_id %in% tss_temp$transcript_id]
save(tss_u, file="./rdata/hg19_isolated_tss.RData")
```
### Plot Human TBP
```{r tbp, eval=T}
tss <- get(load("./rdata/hg19_isolated_tss.RData"))
pugh_tbp <- readRDS("/data/public_data/pugh_tbp/tbp1_k562_pugh.cl.rds")
zeitlinger_tbp1 <- list(pos=import("/data/public_data/zeitlinger_tbp/GSM1333891_hsap_k562_tbp_chipnexus_01_positive.bw", asRle = T),
neg=import("/data/public_data/zeitlinger_tbp/GSM1333891_hsap_k562_tbp_chipnexus_01_negative.bw", asRle = T))
zeitlinger_tbp2 <- list(pos=import("/data/public_data/zeitlinger_tbp/GSM1333892_hsap_k562_tbp_chipnexus_02_positive.bw", asRle = T),
neg=import("/data/public_data/zeitlinger_tbp/GSM1333892_hsap_k562_tbp_chipnexus_02_negative.bw", asRle = T))
normalize_and_combine <- function(rep1, rep2=NULL){
rep1_total <- sum(abs(rep1$pos), abs(rep1$neg))
rep1_pos <- rep1$pos / rep1_total * 1000000
rep1_neg <- abs(rep1$neg) / rep1_total * (-1) * 1000000
if(is.null(rep2)){
bw <- list(pos=rep1_pos, neg=rep1_neg)
}else{
rep2_total <- sum(abs(rep2$pos), abs(rep2$neg))
rep2_pos <- rep2$pos / rep2_total * 1000000
rep2_neg <- abs(rep2$neg) / rep2_total * (-1) * 1000000
bw <- list(pos = rep1_pos + rep2_pos, neg = rep1_neg + rep2_neg)
}
bw
}
zeitlinger_tbp <- normalize_and_combine(zeitlinger_tbp1, zeitlinger_tbp2)
pugh_tbp <- normalize_and_combine(pugh_tbp)
tss$tbp <- resize(tss, 201, "center") %>% nexus_regionSums(., zeitlinger_tbp)
high_tss <- tss[order(tss$tbp, decreasing=T)][1:2000]
tbp_pugh_metapeak <- exo_metapeak(high_tss,pugh_tbp, smooth=5, sample_name = "Pugh TBP")
tbp_zeitlinger_metapeak <- exo_metapeak(high_tss, zeitlinger_tbp, smooth=5, sample_name = "Zeitlinger TBP")
localMaxima <- function(x) {
# Use -Inf instead if x is numeric (non-integer)
y <- diff(c(-.Machine$integer.max, x)) > 0L
rle(y)$lengths
y <- cumsum(rle(y)$lengths)
y <- y[seq.int(1L, length(y), 2L)]
if (x[[1]] == x[[2]]) {
y <- y[-1]
}
y
}
plot_exo_metapeak <- function(metapeak, pos.col, neg.col, name){
ymax <- max(abs(metapeak$reads))
peaks <- localMaxima(abs(metapeak$reads))
metapeak$position <- NA
metapeak$position[peaks] <- metapeak$reads[peaks]
metapeak.p <- subset(metapeak, strand == "+")
metapeak.n <- subset(metapeak, strand == "-")
metapeak.p$position <- ifelse(abs(metapeak.p$reads) < quantile(abs(metapeak.p$reads), 0.7), NA, metapeak.p$position)
metapeak.n$position <- ifelse(abs(metapeak.n$reads) < quantile(abs(metapeak.n$reads), 0.7), NA, metapeak.n$position)
metapeak <- rbind(metapeak.p, metapeak.n)
metapeak$label_position <- metapeak$position
metapeak$label_position <- ifelse(metapeak$strand == "+", metapeak$label_position + ymax/20, metapeak$label_position - ymax / 20)
metapeak$label <- NA
metapeak$label[peaks] <- metapeak$tss_distance[peaks]
metapeak$label <- ifelse(abs(metapeak$reads) < quantile(abs(metapeak$reads), 0.1), NA, metapeak$label)
x <- ggplot(metapeak, aes(x=tss_distance, y=reads, fill=strand)) +
geom_area(position="identity") +
scale_fill_manual(values=c(pos.col, neg.col)) +
ggtitle(name) +
theme(panel.grid.major = element_blank(), panel.grid.minor = element_blank(),
panel.background = element_blank(), axis.line.x = element_line(colour = "black"),
axis.line.y = element_line(colour = "black")) +
xlab("distance from TSS (bp)") + ylab("average RPM") +
ylim(-1 * (ymax + ymax / 10), ymax+ ymax/10) +
geom_point(data=metapeak, aes(x=tss_distance, y=position), size=0.5)+
geom_text(data=metapeak, aes(x=tss_distance, y=label_position, label=label))
x
}
plot_exo_metapeak(tbp_pugh_metapeak, "#C14951", "#D67D80", "Pugh TBP")
plot_exo_metapeak(tbp_zeitlinger_metapeak, "#C14951", "#D67D80", "Zeitlinger TBP")
```
```{r}
sessionInfo()
```