-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathGEO_database_and_clusterProfiler.R
159 lines (140 loc) · 5.83 KB
/
GEO_database_and_clusterProfiler.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
rm(list=ls())
setwd('C:\\Users\\R_study\\clusterprofilers')
### GEO download
rm(list=ls())
suppressPackageStartupMessages(library(GEOquery)) ###ignoring tips
library(GEOquery)
gset <- getGEO("GSE42872", GSEMatrix =T, AnnotGPL=F,getGPL=F,destdir='.' ) ### AnnotGPL=F taking no annotation
group<-pData(gset[[1]]) ###garned sample information from list element [[1]]
library(stringr)
group1<-group[,1:2]
group1$type<-rep(c('Control','Vemurafenib'),each=3)
colnames(group1)[2]<-'variable' ### modified specified column names
### read.table read file with suffix .gz
test1<-read.table('./GSE42872_series_matrix.txt.gz',sep = '\t',comment.char = '!',
quote = '',header = T,fill = T)
a1=exprs(gset[[1]]) #obtainned matrix of expression
rownames(test1)<-test1[,1] ### tranfer first column into ID ,in general, pure digital numbers
test1<-test1[,-1]
#### probe transformation
#### BiocManager::install('pd.hugene.1.0.st.v1')
#### BiocManager::install('hugene10sttranscriptcluster')
suppressPackageStartupMessages(library(hugene10sttranscriptcluster.db)) ### the packages
ls("package:hugene10sttranscriptcluster.db") ### all including check
#str(hugene10sttranscriptcluster.db)
###
probe_symbols=toTable(hugene10sttranscriptclusterSYMBOL) ###get corresponding between symbol and probe ID
####then
table(is.element(rownames(test1),probe_symbols$probe_id))
####filter same probe genes
test1<-test1[is.element(rownames(test1),probe_symbols$probe_id),]
###classified by symbol filtering same probe genes
tmp<-by(test1,probe_symbols$symbol,function(x) rownames(x)[which.max(rowMeans(x))]) ###requiring max genes
probes<-as.character(tmp)
##############################
##############################
### filter again
exprSet<-test1[rownames(test1)%in%probes,]
### plotting barplot through ggplot
library(ggplot2)
library(reshape2)
exprSet_m<-exprSet
exprSet_m$ID<-rownames(exprSet)
exprSet_m<-melt(exprSet_m,id.vars = c('ID'))
exprSet_m$variable<-gsub('X.','',exprSet_m$variable)
exprSet_m$variable<-gsub('[.]','',exprSet_m$variable)
exprSet_m<-merge(exprSet_m,group1,by=c('variable'),all=T)
p<-ggplot(exprSet_m)+geom_boxplot(aes(x=variable,y=value,fill=type))
####got sample evolutionary relationship
colnames(exprSet)<-substr(colnames(exprSet),3,12)
colnames(exprSet)<-paste(colnames(exprSet),1:6,sep='_')
hc<-hclust(dist(t(exprSet)))
plot(hc)
#### pca analysis
library(ggfortify)
df<-as.data.frame(t(exprSet))
df$group<-group1$type
autoplot(prcomp(df[,1:ncol(df)-1]),data=df,colour='group')
#### deg analysis，limma
group_list<-as.character(group1$type)
suppressPackageStartupMessages(library(limma))
design<-model.matrix(~0+factor(group_list)) ###output design matrix
colnames(design)<-levels(factor(group_list))
rownames(design)<-colnames(exprSet)
#### contrast matrix
contrast_matrix<-makeContrasts(paste0(unique(group_list),collapse = '-'),
levels=design)
#### start
#step1
fit<-lmFit(exprSet,design) ### expression and groupping matrix
#step2
fit1<-contrasts.fit(fit,contrast_matrix)
fit1<-eBayes(fit1) ### default no trend
#step3
tempout<-topTable(fit1,coef=1,n=Inf) ###all results
#### following heatmap and enrichment analysis
#heatmap
library(pheatmap)
topgene<-head(rownames(tempout),n=30)
topgene_expr<-exprSet[topgene,]
topgene_expr=t(scale(t(topgene_expr))) ###scaling
topgene_expr<-log10(topgene_expr)
topgene_expr[which(is.nan(topgene_expr))]<-0
pheatmap(topgene_expr)
####
plot(tempout$logFC,-log10(tempout$P.Value))
tempout$regulate<-NA
library(ggplot2)
for (i in 1:nrow(tempout) ) {
if ( as.double(tempout[i,ncol(tempout)-3]) <=0.05 && as.double(tempout[i,1]) >=2 ) {
tempout[i,ncol(tempout)] <- "up"
} else if ( as.double(tempout[i,ncol(tempout)-3]) <= 0.05 && as.double(tempout[i,1]) <=0-2 ) {
tempout[i,ncol(tempout)] <- "down"
} else {
tempout[i,ncol(tempout)] <- "normal"
}
}
p<-ggplot(tempout)+geom_point(aes(x=logFC,y=-log10(P.Value),color=regulate),alpha=0.4,size=1.75)
p<-p+theme_set(theme_bw(base_size = 20))+xlab('log2 fold change')+ylab('-log10 P_Value')
p<-p+ggtitle(c('volcano of all table'))+theme(plot.title = element_text(size=15,hjust = 0.5))
p<-p+scale_colour_manual(values=c('blue','black','red'))
#### hypergeometric test for enrichment
suppressPackageStartupMessages(library(clusterProfiler))
#### top 1000 genes to tranfer symbol into entrez ID
#### merge symbol ID
colnames(probe_symbols)[1]<-'ID'
tempout$ID<-NULL ###remove last column
tempout$ID<-rownames(tempout)
tempout1<-merge(tempout,probe_symbols,by='ID',all=T)
tempout2<-tempout1[complete.cases(tempout1),] ###abandon NA columns
### sort
tempsort<-tempout2[unlist(sort(tempout2$regulate,na.last=NA,decreasing = FALSE,index.return=TRUE)[2]),]
rownames(tempsort)<-tempsort$symbol
tempsort1<-tempsort[,-c(1,ncol(tempout1))]
gene<-head(rownames(tempsort1),1000)
#### bitr function working on tranfer symbol into entrezID
tran_gene_to_entrez<-bitr(gene,fromType = 'SYMBOL',toType = c('ENSEMBL','ENTREZID'),OrgDb ='hugene10sttranscriptcluster.db')
####start kegg enrichment,need more miniutes
kk<-enrichKEGG(gene=tran_gene_to_entrez$ENTREZID,
organism = 'hsa',
pvalueCutoff =0.05)
head(kk)[,1:6]
#### plotting
data(geneList,package = 'DOSE') ###load data
View(geneList) ###fold change， rownames' entrez ID
#### tranferring entrez iD
symbol_id<-rownames(tempsort1)
symbol2entrezID<-bitr(symbol_id,fromType = 'SYMBOL',toType = c('ENSEMBL','ENTREZID'),OrgDb ='hugene10sttranscriptcluster.db')
combined<-merge(symbol2entrezID,tempsort,by.x='SYMBOL',by.y='symbol')
combined1<-combined$logFC
names(combined1)<-combined$ENTREZID
View(combined1)
combined1<-sort(combined1,decreasing = T) ###sort mandatory
####
kk2<-gseKEGG(geneList = combined1,
organism='hsa',
nPerm=1000,
minGSSize = 120,
pvalueCutoff = 0.05,
verbose=FALSE)
gseaplot(kk2,geneSetID = 'hsa04145')