RNA-seq(10):KEGG通路可视化:gage和pathview
这部分直接从上部分RNA-seq(9):富集分析(功能注释)的数据而来，当然如果你上部分数据存盘了，这部分直接导入并进行转换就可以。这里我们先用另外一个R包 gage package (Generally Applicable Gene-set Enrichment for Pathway Analysis)进行KEGG 富集分析，这样也可以和上部分进行比较。
提前说明几个问题
- kegg的物种缩写在这里查看
- 我们使用 gage package (Generally Applicable Gene-set Enrichment for Pathway Analysis) 进行通路分析。点击下载 gage package workflow vignette for RNA-seq pathway analysis查看gage包工作流程。一旦有了富集的通路list，就可以使用pathview 包进行通路可视化。当然这会用到上下调信息。
- 用pathview进行可视化
先安装R包
source("https://bioconductor.org/biocLite.R")
biocLite("gage")
biocLite("pathview")
biocLite("gageData")
library("pathview")
library("gage")
library("gageData")
install.packages("dplyr")
library("dplyr")
#library(clusterProfiler)
#library(DOSE)
#library(stringr)
#library(org.Mm.eg.db)加载数据
data(kegg.sets.mm)
data(sigmet.idx.mm)
kegg.sets.mm = kegg.sets.mm[sigmet.idx.mm]
head(kegg.sets.mm,3)
setwd("F:/rna_seq/data/matrix")
sig.gene<-read.csv(file="DEG_treat_vs_control.csv")
gene.df<-bitr(gene, fromType = "ENSEMBL",
toType = c("SYMBOL","ENTREZID"),
OrgDb = org.Mm.eg.db)
head(sig.gene)> head(sig.gene)
X baseMean log2FoldChange lfcSE stat pvalue padj
1 ENSMUSG00000003309 548.1926 3.231611 0.2658125 12.157485 5.234568e-34 8.193146e-30
2 ENSMUSG00000046323 404.1894 3.067050 0.2628220 11.669687 1.820923e-31 1.425055e-27
3 ENSMUSG00000001123 341.8542 2.797485 0.2766499 10.112004 4.887441e-24 2.549941e-20
4 ENSMUSG00000023906 951.9460 2.382307 0.2510718 9.488551 2.342684e-21 9.116395e-18
5 ENSMUSG00000018569 485.4839 3.136031 0.3312999 9.465836 2.912214e-21 9.116395e-18
6 ENSMUSG00000000184 601.0842 -2.827750 0.3154171 -8.965112 3.099648e-19 8.085948e-16开始用gage包进行富集分析，gage()函数需要fold change 和Entrez gene IDs
foldchanges = sig.gene$log2FoldChange
names(foldchanges)= gene.df$ENTREZID
head(foldchanges)如下显示：
> head(foldchanges)
11768 73708 16859 54419 53624 12444
3.231611 3.067050 2.797485 2.382307 3.136031 -2.827750开始pathway分析，获取结果
keggres = gage(foldchanges, gsets = kegg.sets.mm, same.dir = TRUE)
# Look at both up (greater), down (less), and statatistics.
lapply(keggres, head)显示为
> lapply(keggres, head)
$greater
p.geomean stat.mean p.val q.val set.size exp1
mmu04514 Cell adhesion molecules (CAMs) 0.2680462 0.6286461 0.2680462 0.5360924 12 0.2680462
mmu04510 Focal adhesion 0.6382502 -0.3594187 0.6382502 0.6382502 10 0.6382502
mmu04144 Endocytosis NA NaN NA NA 8 NA
mmu03008 Ribosome biogenesis in eukaryotes NA NaN NA NA 0 NA
mmu04141 Protein processing in endoplasmic reticulum NA NaN NA NA 0 NA
mmu04740 Olfactory transduction NA NaN NA NA 1 NA
$less
p.geomean stat.mean p.val q.val set.size exp1
mmu04510 Focal adhesion 0.3617498 -0.3594187 0.3617498 0.7234996 10 0.3617498
mmu04514 Cell adhesion molecules (CAMs) 0.7319538 0.6286461 0.7319538 0.7319538 12 0.7319538
mmu04144 Endocytosis NA NaN NA NA 8 NA
mmu03008 Ribosome biogenesis in eukaryotes NA NaN NA NA 0 NA
mmu04141 Protein processing in endoplasmic reticulum NA NaN NA NA 0 NA
mmu04740 Olfactory transduction NA NaN NA NA 1 NA
$stats
stat.mean exp1
mmu04514 Cell adhesion molecules (CAMs) 0.6286461 0.6286461
mmu04510 Focal adhesion -0.3594187 -0.3594187
mmu04144 Endocytosis NaN NA
mmu03008 Ribosome biogenesis in eukaryotes NaN NA
mmu04141 Protein processing in endoplasmic reticulum NaN NA
mmu04740 Olfactory transduction NaN NA得到pathway
keggrespathways = data.frame(id=rownames(keggres$greater), keggres$greater) %>%
tbl_df() %>%
filter(row_number()<=10) %>%
.$id %>%
as.character()
keggrespathways结果如下：
> keggrespathways
[1] "mmu04514 Cell adhesion molecules (CAMs)" "mmu04510 Focal adhesion"
[3] "mmu04144 Endocytosis" "mmu03008 Ribosome biogenesis in eukaryotes"
[5] "mmu04141 Protein processing in endoplasmic reticulum" "mmu04740 Olfactory transduction"
[7] "mmu03010 Ribosome" "mmu04622 RIG-I-like receptor signaling pathway"
[9] "mmu04744 Phototransduction" "mmu04062 Chemokine signaling pathway" # Get the IDs.
keggresids = substr(keggrespathways, start=1, stop=8)
keggresids> keggresids
[1] "mmu04514" "mmu04510" "mmu04144" "mmu03008" "mmu04141" "mmu04740" "mmu03010" "mmu04622" "mmu04744" "mmu04062"最后，可以通过pathview包中的pathway()函数画图。下面写一个函数，这样好循环画出上面产生的前10个通路图。
# 先定义画图函数
plot_pathway = function(pid) pathview(gene.data=foldchanges, pathway.id=pid, species="mmu", new.signature=FALSE)
# 同时画多个pathways，这些plots自动存到工作目录
tmp = sapply(keggresids, function(pid) pathview(gene.data=foldchanges, pathway.id=pid, species="mmu"))显示如下
> tmp = sapply(keggresids, function(pid) pathview(gene.data=foldchanges, pathway.id=pid, species="mmu"))
Info: Downloading xml files for mmu04514, 1/1 pathways..
Info: Downloading png files for mmu04514, 1/1 pathways..
'select()' returned 1:1 mapping between keys and columns
Info: Working in directory F:/rna_seq/data/matrix
Info: Writing image file mmu04514.pathview.png
Info: Downloading xml files for mmu04510, 1/1 pathways..
Info: Downloading png files for mmu04510, 1/1 pathways..
'select()' returned 1:1 mapping between keys and columns然后我们去工作目录，查看KEGG pathway，我放三张图查看下：
mmu04144.pathview.png
mmu04510.pathview.png
mmu04514.pathview.png
至此，KEGG 通路可视化完成
后记：
更详细的可视化见（可以从counts开始）
本文参与 腾讯云自媒体分享计划，分享自作者个人站点/博客。
原始发表：2018.08.04 ，如有侵权请联系 cloudcommunity@tencent.com 删除
评论
登录后参与评论
推荐阅读
目录