数据挖掘—GPL5175探针注释处理

rm(list = ls())
#GPL5175探针处理办法
#获取注释文件的方法一
library(GEOquery)
gpl_number='GPL5175'
# a = getGEO(gpl_number,destdir = ".")
# b = a@dataTable@table
# colnames(b)

#获取注释文件的方法二
library(tinyarray)
get_gpl_txt(gpl_number) 
#获取表格文件的下载链接
#接下来是复制网址去浏览器下载、放在工作目录下、读取、提取探针id和基因symbol(没有现成的需要拆分和转换)，不同文件代码不统一，等看同学们的例子。
#注意:最终的数据ids只能有两列，第一列列名是probe_id,第二列列名是symbol,且都是字符型，否则后面代码要报错咯。
tmp <- data.table::fread("GPL5175.txt",header = T,data.table = F)
extract_first_symbol <- function(gene_assignment) {
  if (is.na(gene_assignment) || gene_assignment == "---") {
    return(NA)
  }
  # 找到第一个" // "的位置
  first_delim <- regexpr(" // ", gene_assignment, fixed = TRUE)
  if (first_delim == -1) return(NA)
  # 找到第二个" // "的位置 
  rest_string <- substr(gene_assignment, first_delim + 4, nchar(gene_assignment))
  second_delim_in_rest <- regexpr(" // ", rest_string, fixed = TRUE)
  if (second_delim_in_rest == -1) return(NA)
  # 提取之间的内容
  symbol <- substr(rest_string, 1, second_delim_in_rest - 1)
  return(symbol)
}

tmp$symbol <- sapply( tmp$gene_assignment, extract_first_symbol)
unique_symbols_no_na <- unique( tmp$symbol[!is.na( tmp$symbol)])

ids <-  tmp[, c("ID", "symbol")]  
head(ids)
colnames(ids) <- c("probe_id", "symbol")
# 去除空值和重复
ids <- ids[!is.na(ids$symbol) & ids$symbol != "", ]
ids$probe_id <- as.character(ids$probe_id)
head(ids)
# probe_id       symbol
# 1   2315100      DDX11L2
# 11  2315125       OR4F17
# 15  2315147 LOC100288692
# 16  2315160     FLJ45445
# 17  2315163 LOC100132062
# 19  2315191 LOC100507658
str(ids)
# 'data.frame':	33475 obs. of  2 variables:
#   $ probe_id: chr  "2315100" "2315125" "2315147" "2315160" ...
# $ symbol  : chr  "DDX11L2" "OR4F17" "LOC100288692" "FLJ45445" ...
save(ids,file = "step0_ids.Rdata")