add sub-category information of the KEGG pathways

[GuangchuangYu]

KEGG can be divided into 7 categories, see https://www.genome.jp/kegg/pathway.html.

It is easy to incorporate this information in the enrichKEGG() and gseaKEGG() results, so that we can use this information to filter the results or to differentiate the pathways for visualization.

reference: https://mp.weixin.qq.com/s/17ujVhcrkX1DLsUJBtUGEw.

[Potato-tudou]

A function embedding KEGG api (https://rest.kegg.jp/get/) could be used to get the classification info of a KEGG term with the given ID. However, the purpose of this function is to keep the classification info updated, it is effective, but not so high-efficient. Still, handle things locally is more elegant.

library(pacman)
pacman::p_load(httr, jsonlite, magrittr)

class_t <- function(term) {
    res <- paste( "https://rest.kegg.jp/get/",
                  term, sep = '') %>% GET()
    res_info <- res$content %>% rawToChar()
    class_res <- gsub('^.*CLASS\\s*|\\s*PATHWAY_MAP.*$', '', res_info)
    return(class_res)
  }
class_t("mmu04380")
[1] "Organismal Systems; Development and regeneration"

#Then the user can easily get all the classification of KEGG terms in enrichKEGG result by: 
lapply(kegg_res@result$ID, class_t)

[Potato-tudou]

Now the category of a certain kegg term can be extracted by the help of the referred url ("https://pathview.uncc.edu/data/khier.tsv").

'''
k.info <- read.table("https://pathview.uncc.edu/data/khier.tsv", header = T) %>%
separate(pathway, c("ID","Description"), extra = "merge",fill = "right")

getKEGG_cat <- function(ID, k_info) {
cleanID <- function (id_num) {
gsub("[a-z]", "", id_num)
}
inputID <- cleanID(ID)
k_info[(k_info$ID == inputID),]$category
}

getKEGG_cat(ID = "mmu04380", k_info = k.info)
[1] "Organismal Systems"

'''

[GuangchuangYu]

@Potato-tudou pls learn how to format your code first.

Refer to point 2 mentioned by Yonghe, #1 (comment).

[Potato-tudou]

k.info <- read.table("https://pathview.uncc.edu/data/khier.tsv", header = T) %>%
separate(pathway, c("ID","Description"), extra = "merge",fill = "right")

getKEGG_cat <- function(ID, k_info) {
cleanID <- function (id_num) {
gsub("[a-z]", "", id_num)
}
inputID <- cleanID(ID)
k_info[(k_info$ID == inputID),]$category
}

getKEGG_cat(ID = "mmu04380", k_info = k.info)

[Potato-tudou]

I think it's better to use the result of enrichKEGG() as an input. So here it is:

listKEGG_cat <- function (enrich_res) {
  k.info <- read.table("https://pathview.uncc.edu/data/khier.tsv", header = T) %>% 
    separate(pathway, c("ID","Description"), extra = "merge",fill = "right")
  getKEGG_cat <- function(ID, k_info) {
    cleanID <- function (id_num) {
      gsub("[a-z]", "", id_num)
    }
    inputID <- cleanID(ID)
    k_info[(k_info$ID == inputID),]$category
  }
  lapply(enrich_res@result$ID,getKEGG_cat, k_info = k.info) %>% unlist()
}

[GuangchuangYu]

see also YuLab-SMU/clusterProfiler#236.