GEO表达芯片平台 — GPL14951，注释文件探索过程

#根据注释包的帮助文档，探索了下illuminaHumanv4.db
library(illuminaHumanv4.db)
x <- illuminaHumanv4ENTREZID
mapped_probes <- mappedkeys(x)
head(mapped_probes, 10)
> head(mapped_probes, 10)
 [1] "ILMN_1343291" "ILMN_1343295" "ILMN_1651209" "ILMN_1651221" "ILMN_1651228"
 [6] "ILMN_1651229" "ILMN_1651230" "ILMN_1651232" "ILMN_1651235" "ILMN_1651237"

anno <- data.table::fread("GPL14951-11332.txt")#读取
colnames(anno)
probe2symbol <- anno[,c("ID","Symbol")]#取需要的列
colnames(probe2symbol) <- c("PROBE_ID", "SYMBOL_ID")#改名，让他适合下面的自定义函数
{
  p2g <- function(eset,probe2symbol){
    library(dplyr)
    library(tibble)
    library(tidyr)
    eset <- as.data.frame(eset)
    p2g_eset <- eset %>% 
      rownames_to_column(var="PROBE_ID") %>% #合并探针的信息
      inner_join(probe2symbol,by="PROBE_ID") %>% #去掉多余信息
      select(-PROBE_ID) %>% #重新排列
      dplyr::select(SYMBOL_ID,everything()) %>% #求出平均数(这边的点号代表上一步产出的数据)
      mutate(rowMean = rowMeans(.[grep("GSM", names(.))])) %>% #去除symbol中的NA
      filter(SYMBOL_ID != "NA") %>% #把表达量的平均值按从大到小排序
      arrange(desc(rowMean)) %>% # symbol留下第一个
      distinct(SYMBOL_ID,.keep_all = T) %>% #反向选择去除rowMean这一列
      dplyr::select(-rowMean) %>% # 列名变成行名
      column_to_rownames(var = "SYMBOL_ID")
    save(p2g_eset, file = "p2g_eset.Rdata")
    return(p2g_eset)
  }
  p2g_eset <- p2g(eset = eset, probe2symbol = probe2symbol)
  load("p2g_eset.Rdata")
}