KEGG通路的从属/注释信息如何获取

富集分析经常做，之前只知道GO term有从上到下的层次关系，今天才知道KEGG pathway也有类似的分层关系。

起因是我准备更新一下自己的kegg富集结果展示图，之前一直画的这种图，略显朴素了。

正好搜到了这张图，还挺好看的（主要是配色很清新）

https://doi.org/10.1016/j.phymed.2021.153714

这张图并不难（等有空把代码整理出来发给大家），可是这个kegg pathway annotation我真没见过呀！去官网看了看，能找到，但没法下载

https://www.genome.jp/kegg/pathway.html

在朋友圈求助后，费了很大劲才把这个问题解决。中间好几个小伙伴都很热心地给了我帮助，感谢他们！

以下是我整理的KEGG pathway annotation文件的网盘链接，用到的代码和数据也在里面，想改善一下kegg富集图的朋友可以看看。

链接：https://pan.baidu.com/s/18pwYZGGZSuk2_LGW8_nQFQ
提取码：ihyn
(PS: 觉得有用可以给我本期推送的第1篇推文点个赞呀，谢谢了！)

最终能得到这个表格

整理这个文件的思路如下

1. 浏览器打开这个网页(https://www.genome.jp/kegg/pathway.html)，然后查看网页源代码（一般是鼠标右键），就能看到这个：

先不要被它吓到，后面的信息提取都是基于这些字符串

2. 然后复制粘贴到一个文本文件kegg_html.txt，删掉前188行左右，后17行左右（删掉的这些行明显不含有用信息），得到kegg_html_copy.txt。

3. 然后运行我编写的代码pre.R就能得到最终表格了

library(tidyverse)
tmp1=readLines("kegg_html_copy.txt")

tmp2=c()
for (li in 1:length(tmp1)) {
  if(str_detect(tmp1[li],">[0-9]")) {
    tmp2=append(tmp2,tmp1[li])
  }
}

###
big_anno=""
small_anno=""
final_lines=c()
for (li in 1:length(tmp2)) {
  if (str_detect(tmp2[li],">[0-9]//. ")) {
    tmp_anno=str_extract(tmp2[li],">.*<")
    tmp_anno=str_split(tmp_anno,"//. ")[[1]][2]
    tmp_anno=str_split(tmp_anno,"<")[[1]][1]
    big_anno=tmp_anno
  } else if (str_detect(tmp2[li],">[0-9]//.[0-9]{1,2} ")) {
    tmp_anno2=str_extract(tmp2[li],">.*<")
    tmp_anno2=str_replace(tmp_anno2,"^>[0-9]//.[0-9]{1,2} ","")
    tmp_anno2=str_split(tmp_anno2,"<")[[1]][1]
    small_anno=tmp_anno2
  } else if (str_detect(tmp2[li],">[0-9]{5}")){
    element1=str_extract(tmp2[li],">[0-9]{5}")
    element1=str_split(element1,">")[[1]][2]
    
    if (!str_detect(tmp2[li],"hsa//+pathogen")) {
      element2=str_extract(tmp2[li],"pathway///[a-zA-Z]{2,4}[0-9]{5}")
      element2=str_split(element2,"///")[[1]][2]
      
      element3=str_extract(tmp2[li],"pathway.*?<") #非贪婪匹配
      element3=str_extract(element3,">.*<")
      element3=str_split(element3,">")[[1]][2]
      element3=str_split(element3,"<")[[1]][1]
    } else {
      element2="organism:hsa+pathogen"
      element3=str_extract(tmp2[li],"hsa//+pathogen.*?<") #非贪婪匹配
      element3=str_extract(element3,">.*<")
      element3=str_split(element3,">")[[1]][2]
      element3=str_split(element3,"<")[[1]][1]
    }
    
    tmp_line=paste(as.character(element1),element2,element3,big_anno,small_anno,sep = ";")
    final_lines=append(final_lines,tmp_line)
  }else{
    print(tmp2[li])
  }
}

###
final_df=as.data.frame(final_lines)
colnames(final_df)="V1"
final_df=final_df%>%apply(1, function(x){as.data.frame(str_split(x,";")[[1]])})
final_df=as.data.frame(final_df)
final_df=as.data.frame(t(final_df))
rownames(final_df)=NULL
colnames(final_df)=NULL
colnames(final_df)=c("ID","Pathway Identifier","Pathway","big annotion","small annotion")

###
library(xlsx)
write.xlsx(final_df,file = "kegg_info.xlsx",col.names = T,row.names = F)