-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathRNAseq_CohesinAML_adultHSCs_GEXplots.Rmd
253 lines (236 loc) · 10.9 KB
/
RNAseq_CohesinAML_adultHSCs_GEXplots.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
---
title: "RNAseq Cohesin AML - adult HSCs - Expression of Cohesin genes"
author: "Alexander Fischer"
date: "05 2022"
output: html_document
---
#rbioc_3-12
# RNAseq comparison Cohesin AML - adult HSCs
# Loading libraries and data
## libraries
```{r, echo = FALSE, include = FALSE}
library(edgeR)
library(GGally)
library(ggplot2)
library(ggrepel)
library(gplots)
library(RColorBrewer)
library(reshape2)
library(sqldf)
library(ggpubr)
library(Rtsne)
library(dplyr)
library(readr)
library(tidyr)
library(stringr)
library(gridExtra)
library(cowplot)
```
## Defining path variables at the start
```{r}
DIR_DATA="/misc/data"
PROJDIR="/misc/data/analysis/project_cohesin/Cohesin_AML"
WORKDIR=file.path(PROJDIR,"RNAseq_comparison_HSCs")
METADIR=file.path(WORKDIR,"Metadata") #input direcotry for respective metadata file
RPGDIRHSC=file.path(DIR_DATA,"processedData/mapping/RNA/GRCh38/RNAseq/CD34/adult_CD34")
METADIRAML=file.path(PROJDIR,"RNAseq/Metadata")
ANALDIRAML=file.path(PROJDIR,"RNAseq/Analysis/Resulttables/RNAseq_Cohesin_AML_AF3")
dir.create(file.path(WORKDIR,"Analysis"))
ANALDIR=file.path(WORKDIR,"Analysis/Resulttables") #output of analysis results
dir.create(file.path(ANALDIR))
FIGDIR=file.path(WORKDIR,"Analysis/Plots") #output of figure directories
dir.create(file.path(FIGDIR))
STIDp="/misc/software/ngs/genome/sequence/GRCh38.PRI_p10/STAR_transcriptIDshort.txt" #this is the path to the ShortTranscriptID for a fully functional annotation of the RCT
```
## create a genes dataframe
```{r}
stid<-read.delim(STIDp,skip = 4,header = FALSE)[,1]
genes.df<-as.data.frame(strsplit2(stid,"[$]"))
colnames(genes.df)<-c("EnsemblID","GeneSymbol","Length","GeneType")
genes.df$EnsemblID<-as.character(genes.df$EnsemblID)
genes.df$GeneSymbol<-as.character(genes.df$GeneSymbol)
genes.df$Length<-as.numeric(as.character(genes.df$Length))
genes.df$GeneType<-as.character(genes.df$GeneType)
head(genes.df)
```
## get metadata and add a cell type column
```{r}
metad_AMLcomplete<-read.table(file.path(METADIRAML, "RNAseq_Metadata_AML_STAG2_RAD21.txt"),sep = "\t",header=TRUE)
metad_AML<-metad_AMLcomplete[,c("Sample_Name","ID","group","sex")]
metad_AML$Celltype<-"AML"
metad_adultHSCs<-read.csv(file.path(METADIR,"metad_adultHSC_simple.csv"))
colnames(metad_adultHSCs)<-c("Sample_Name","ID","group","sex")
metad_adultHSCs$Celltype<-"adult_HSCs"
```
## for adult HSPCs a new RCT has to be created
```{r}
rpgl<-list.files(path = RPGDIR,pattern = "ReadsPerGene.txt",ignore.case = FALSE)
sortvec<-sapply(as.character(metad_adultHSCs$Sample_Name), function(x) grep(x,rpgl))
counts<-as.data.frame(matrix(),row.names = NULL)
for (i in sortvec) {
rpgf<-read.delim(file.path(RPGDIR,rpgl[i]),sep = "\t",header = FALSE,check.names = TRUE,skip = 4,row.names = NULL)
counts<-cbind(counts,rpgf[,4])
}
adultHSCs_RCT<-counts[,(-1)]
colnames(adultHSCs_RCT)<-metad_adultHSCs$Sample_Name
rownames(adultHSCs_RCT)<-stid
ncol(adultHSCs_RCT) #all samples there?
head(adultHSCs_RCT)
write.table(adultHSCs_RCT,file=file.path(ANALDIR,"RCT_adultHSCs.txt"),sep = "\t",col.names=TRUE, quote=FALSE)
```
## Read in pre-existing Read-count tables
```{r}
AML_RCT<-read.table(file.path(PROJDIR,"RNAseq","RNAseq_AML_STAG2_RAD21_counts.raw.txt"),header=TRUE)
adultHSCs_RCT<-read.table(file=file.path(ANALDIR,"RCT_adultHSCs.txt"),sep = "\t",header=TRUE)
```
## read in pre-existing statistics from individ. AML comparisons qlf tests
```{r}
listSA2mut2<-read.table(file=file.path(ANALDIRAML,"STAG2pat_vs_CTRL","qstat_STAG2.vs.CTRL.glm.txt"))
listRAD21mut2<-read.table(file=file.path(ANALDIRAML,"RAD21pat_vs_CTRL","qstat_RAD21.vs.CTRL.glm.txt"))
```
# prepare data
## merge Read-count tables and metadata respecitively
```{r}
#merge metadata #AML + KD
Metamerged<-rbind(metad_AML,metad_adultHSCs)
###add labelID column
Metamerged$label<-paste0(Metamerged$group,"_",Metamerged$ID)
#merge RCT tables
RCTmerged<-cbind(AML_RCT,adultHSCs_RCT)
write.table(RCTmerged,file=file.path(ANALDIR,"RNAseq_AML_CTRL_STAG2mut_RAD21mut_adultHSCs.counts.raw.txt"),sep = "\t",col.names=TRUE,quote=FALSE)
```
# generate dgelist object and norm. counts
```{r}
#DGEList and filtering
dgel_AMLHSCad <- DGEList(counts = RCTmerged, group = Metamerged$group, genes = genes.df)
keep <- rowSums(cpm(dgel_AMLHSCad)>1) >= 12
dgel_AMLHSCad <- dgel_AMLHSCad[keep, , keep.lib.sizes=FALSE]
dgel_AMLHSCad <- calcNormFactors(dgel_AMLHSCad)
summary(keep)
#cpm rpkm transformation
d.log.cpm3 <- cpm(dgel_AMLHSCad, prior.count = 2, log = TRUE)
d.log.rpkm3 <- rpkm(dgel_AMLHSCad, prior.count = 2, normalized.lib.sizes = TRUE, log = TRUE)
write.table (d.log.rpkm3, file = file.path(ANALDIR,"RNAseq_AML_CTRL_STAG2mut_RAD21mut_adultHSC.log.rpkm.txt"), sep = "\t", col.names=NA, quote=FALSE)
write.table (d.log.cpm3, file = file.path(ANALDIR,"RNAseq_AML_CTRL_STAG2mut_RAD21mut_adultHSC.log.cpm.txt"), sep = "\t", col.names=NA, quote=FALSE)
#scaled cpms (z score)
d.log.cpm.scaled.transposed3<-(scale(t(d.log.cpm3)))
d.log.cpm.scaled3<-data.matrix(t(d.log.cpm.scaled.transposed3))
```
# Gene expression barplots showing AML groups and adult CD34 samples
## calculate average data
```{r}
###create a table that contains the individual values to show and transform
Individ_pat.rpkm.m<-melt(d.log.rpkm3)
Individ_pat.rpkm.m$group<-rep(Metamerged$group, each=length(d.log.rpkm3[,1]))
##summarize mean + sd
library(plyr)
###stack data
dat.st<-stack(as.data.frame(d.log.rpkm3)) ##stack data to a single column
dat.st$genename<- rep(rownames(d.log.rpkm3), times = ncol(d.log.rpkm3)) ##add genenames as extra col
dat.st$type=rep(Metamerged$group, each=length(d.log.rpkm3[,1])) ##add group identity as extra col
###summarize by mutation group
mean.d.log.rpkm3<-ddply(dat.st, .(type,genename),
summarize,
mean= round(mean(values),2),
sd=round(sd(values),2))
```
## define GEX-BARplot function with optional lables and optional fixed y limits
```{r}
#labels for bars displaying means
collabvec<-c("CTRL\nAML","STAG2\nmut","RAD21\nmut","healthy\nHSCs") #labels to show under bars
collablvecempty<-c(rep("",4)) #if no labels are desired
#function showing individual values as dots and showing significance levels of CTRL vs STAG2mut or CTRL vs RAD21mut
GENEX_func_means<-function(GOI,y1=NA,y2=NA,bottom="",legloc="none",labs="HICpat",col.labs=collabvec,pointsize=rel(2)){
GOIsearchterm<-paste0("\\$",GOI,"\\$")
#individual HIC pat data
GOIdataPoints<-Individ_pat.rpkm.m[grep(GOIsearchterm,Individ_pat.rpkm.m$Var1),]
GOIdataPoints$sd<-0
colnames(GOIdataPoints)<-c("genename","label","value","group","sd")
rownames(GOIdataPoints)<-GOIdataPoints$label
#average data of all patients
GOIdata<-mean.d.log.rpkm3[grep(GOIsearchterm,mean.d.log.rpkm3$genename),]
#GOIdata<-GOIdata[GOIdata$type=="CTRL"|GOIdata$type=="STAG2"|GOIdata$type=="RAD21",] #subset only for the groups of interest
GOIdata$label<-c("CTRL-average","STAG2mut-average","RAD21mut-average","HSC-average")
col_order <- c("genename", "label","mean", "type","sd")
GOIdata2 <- GOIdata[, col_order]
rownames(GOIdata2)<-GOIdata2$label
colnames(GOIdata2)<-c("genename","label","value","group","sd")
GOIdatM<-rbind(GOIdata2,GOIdataPoints) ##merge the two
i<-c(1:nrow(GOIdatM))
#select the HIC patients by patID which should be labelled by geom text
GOIdatM$label2[i]<-sapply(GOIdatM$label[i],function(x) strsplit(as.character(x), "_")[[1]][4])
GOIdatM_HICPAT<-GOIdatM %>% filter(label2 %in% c("2236","1551","1747","3323","3488","6246","5285","UKR186","7314","12557","41580a","12514","12567","24603","2193"))
#check if lables are desired
if (labs=="none"){
GOIdatM_HICPAT$label2<-""
}
#pvalue data
pvaldataSA2<-listSA2mut2[grep(GOIsearchterm,row.names(listSA2mut2)),]
pvalSA2<-pvaldataSA2$FDR
pvaldataRAD21<-listRAD21mut2[grep(GOIsearchterm,row.names(listRAD21mut2)),]
pvalRAD21<-pvaldataRAD21$FDR
##symbol is assigned to significance level for SA2 and RAD21 group
if (pvalSA2 < 0.001) { siglvlSA2 <- "***" } else if (pvalSA2 < 0.01) { siglvlSA2 <- "**" } else if (pvalSA2 < 0.05) { siglvlSA2 <- "*" } else { siglvlSA2 <- "ns" }
if (pvalRAD21 < 0.001) { siglvlRAD21 <- "***" } else if (pvalRAD21 < 0.01) { siglvlRAD21 <- "**" } else if (pvalRAD21 < 0.05) { siglvlRAD21 <- "*" } else { siglvlRAD21 <- "ns" }
##fonstize and face is assigned to significance level for SA2 and RAD21 group
if (pvalSA2 < 0.05){
fsSA2 <- rel(10)
ffSA2 <- "bold"}
else {fsSA2 <- rel(8)
ffSA2 <- "italic"}
if (pvalRAD21 < 0.05){
fsRAD <- rel(10)
fsRAD <- "bold"}
else {fsRAD <- rel(8)
ffRAD <- "italic"}
##y coordinate of signifcance level symbol is calculated for SA2 and RAD21 group
maxval<-max(GOIdatM$value)
siglvlSTAG2coord_y <- (maxval + 0.05*maxval)
siglvlRAD21coord_y <- (maxval + 0.1*maxval)
#ggplot object
ggplot(GOIdatM,aes(x=group,y=value))+
geom_col(data=GOIdatM[1:4,],aes(x=group,y=value, fill=group),alpha=0.8)+
scale_fill_manual(values = c("CTRL" = "firebrick", "STAG2" = "seagreen", "RAD21" = "mediumvioletred","adultHSCs"="darkorange4")) +
scale_x_discrete(labels=col.labs) +
geom_errorbar(data=GOIdatM[1:4,],aes(ymin=value[1:4]-sd[1:4], ymax=value[1:4]+sd[1:4]), width=.2,position=position_dodge(.9)) +
geom_jitter(data = GOIdatM[4:nrow(GOIdatM),], aes(x=group,y=value,fill=group,color=group),alpha=0.6,size=pointsize, width = 0.1)+
scale_color_manual(values = c("CTRL" = "lightsalmon", "STAG2" = "springgreen", "RAD21" = "violet","adultHSCs" = "coral")) +
geom_text_repel(data = GOIdatM_HICPAT,aes(label=label2),
size=rel(2),
nudge_x=0.5,
segment.size=0.3,
min.segment.length=0.3,
point.padding=0,
segment.alpha=0.5,
max.overlaps = 50,
force = 2,
show.legend = FALSE) +
theme(
legend.position=legloc,
axis.text.x = element_text(angle=0, size=rel(2)),
axis.text.y=element_text(size=rel(3.5)),
axis.title=element_text(size=rel(2),face="bold"),
plot.title = element_text(size = rel(3.5), face = "bold.italic",hjust=0.5),
panel.border = element_blank(),
panel.grid.major = element_blank(),
panel.grid.minor = element_blank(),
panel.background = element_blank(),
axis.line = element_line(colour = "black"),) +
coord_cartesian(ylim = c(y1,y2))+
xlab(bottom) + ylab("mean log rpkm") +
ggtitle(GOI) +
annotate("text", x=1.5, y=(siglvlSTAG2coord_y+0.02*siglvlSTAG2coord_y), label= siglvlSA2, size=fsSA2,fontface=ffSA2) +
annotate("text", x=2, y=(siglvlRAD21coord_y+0.02*siglvlRAD21coord_y), label= siglvlRAD21, size=fsRAD,fontface=ffRAD) +
annotate("segment", x = 1, xend = 2, y = siglvlSTAG2coord_y, yend = siglvlSTAG2coord_y)+
annotate("segment", x = 1, xend = 3, y = siglvlRAD21coord_y, yend = siglvlRAD21coord_y)
}
```
## plot GEX for Cohesin components and GOIs identified in previous analyses
```{r}
#cohesin components
CohComp=c("STAG2","STAG1","RAD21","SMC1A","SMC3") ###Cohesin core subunits only
PlotsCohComp = lapply(CohComp, GENEX_func_means,y1=0,y2=6.5,labs="none",pointsize=rel(3))
pdf(file = file.path(FIGDIR,"GEXbarplots","GEX_CohesinCoreComponents_AMLvsHSC.ylim.3.pdf"),height = 10,width = 35)
grid.arrange(grobs=PlotsCohComp,ncol=5)
dev.off()
```