获取更多R语言知识，请关注公众号：医学和生信笔记

    医学和生信笔记 公众号主要分享：1.医学小知识、肛肠科小知识；2.R语言和Python相关的数据分析、可视化、机器学习等；3.生物信息学学习资料和自己的学习笔记！

    从TCGA下载的counts，FPKM和miRNA是很多单个的文件，需要手动合成一个完整的表达矩阵，这是一个高频操作，因此写一段代码简化流程，部分代码来自生信技能树和生信星球。

    把以下代码保存为“get_TCGA_expr.R”，下次直接 source(get_TCGA_expr.R)，就可以使用定义好的3个函数。
    get_TCGAcounts
    get_TCGAfpkm
    get_TCGAmirna

    1. ## 需要提供counts文件所在的文件夹目录和metadata路径
    2. ## 类似于 countsfiles <- "E:/TCGA_COADREAD/mirna/"
    3. ## metadata <- "E:/TCGA_COADREAD/mirna/metadata.cart.2021-05-31.json"
    5. # 获取counts表达矩阵，
    6. get_TCGAcounts <- function(rnaseqpath, metadatapath){
    8.   allfiles <- list.files(path = rnaseqpath ,pattern='*.counts.gz$',recursive=T) 
    10.   mRNA_seq <- lapply(allfiles, function(x){
    11.     result <- data.table::fread(file = file.path(rnaseqpath,x),data.table = F) 
    12.     return( result )
    13.   })
    15.   mRNA_df <- Reduce(cbind, mRNA_seq)
    16.   tem <- nrow(mRNA_df)-5
    17.   mRNA_df <- mRNA_df[1:tem, c(1,seq(2,ncol(mRNA_df),2))]
    19.   save(mRNA_df,file = './TCGA_RNA.Rdata')
    21.   #读取metadata信息
    22.   meta <- jsonlite::fromJSON(metadatapath)
    23.   entity <- meta$associated_entities
    25.   tcga_id <- unlist(lapply(entity,function(x){x["entity_submitter_id"]}))
    26.   file_name <- meta$file_name
    27.   fln2id <- data.frame(filename = file_name, tcgaid = tcga_id)
    29.   #把样本名（tcga_patient_id）变成表达矩阵的列名
    30.   mis2 <- stringr::str_split(allfiles,"/",simplify = T)[,2]
    32.   col_tcga <- fln2id[match(mis2,fln2id$filename),]
    33.   colnames(mRNA_df)[2:ncol(mRNA_df)] <- as.character(col_tcga$tcgaid)
    35.   rnaExpr <<- mRNA_df
    37.   save(rnaExpr,file = "./TCGA_new_RNA.Rdata")
    39.   print(paste("your RNA exprset has been saved at: ", getwd(), sep = ""))
    41. }
    44. # 获取FPKM表达矩阵
    46. get_TCGAfpkm <- function(rnaseqpath, metadatapath){
    48.   allfiles <- list.files(path = rnaseqpath ,pattern='*.FPKM.txt.gz$',recursive=T) 
    50.   mRNA_seq <- lapply(allfiles, function(x){
    51.     result <- data.table::fread(file = file.path(rnaseqpath,x),data.table = F) 
    52.     return( result )
    53.   })
    55.   mRNA_df <- Reduce(cbind, mRNA_seq)
    56.   tem <- nrow(mRNA_df)-5
    57.   mRNA_df <- mRNA_df[1:tem, c(1,seq(2,ncol(mRNA_df),2))]
    59.   save(mRNA_df,file = './TCGA_RNA.Rdata')
    61.   #读取metadata信息
    62.   meta <- jsonlite::fromJSON(metadatapath)
    63.   entity <- meta$associated_entities
    65.   tcga_id <- unlist(lapply(entity,function(x){x["entity_submitter_id"]}))
    66.   file_name <- meta$file_name
    67.   fln2id <- data.frame(filename = file_name, tcgaid = tcga_id)
    69.   #把样本名（tcga_patient_id）变成表达矩阵的列名
    70.   mis2 <- stringr::str_split(allfiles,"/",simplify = T)[,2]
    72.   col_tcga <- fln2id[match(mis2,fln2id$filename),]
    73.   colnames(mRNA_df)[2:ncol(mRNA_df)] <- as.character(col_tcga$tcgaid)
    75.   rnaExpr <<- mRNA_df
    77.   save(rnaExpr,file = "./TCGA_new_RNA.Rdata")
    79.   print(paste("your RNA exprset has been saved at: ", getwd(), sep = ""))
    81. }
    84. # 获取mirna表达矩阵
    86. get_TCGAmirna <- function(rnaseqpath, metadatapath){
    88.   allfiles <- list.files(path = rnaseqpath ,pattern='*.mirnas.quantification.txt$',recursive=T) 
    90.   mRNA_seq <- lapply(allfiles, function(x){
    91.     result <- data.table::fread(file = file.path(rnaseqpath,x),data.table = F) 
    92.     return( result )
    93.   })
    95.   mRNA_df <- Reduce(cbind, mRNA_seq)
    97.   mRNA_df <- mRNA_df[, c(1,seq(2,ncol(mRNA_df),4))]
    99.   save(mRNA_df,file = './TCGA_miRNA.Rdata')
    101.   #读取metadata信息
    102.   meta <- jsonlite::fromJSON(metadatapath)
    103.   entity <- meta$associated_entities
    105.   tcga_id <- unlist(lapply(entity,function(x){x["entity_submitter_id"]}))
    106.   file_name <- meta$file_name
    107.   fln2id <- data.frame(filename = file_name, tcgaid = tcga_id)
    109.   #把样本名（tcga_patient_id）变成表达矩阵的列名
    110.   mis2 <- stringr::str_split(allfiles,"/",simplify = T)[,2]
    112.   col_tcga <- fln2id[match(mis2,fln2id$filename),]
    113.   colnames(mRNA_df)[2:ncol(mRNA_df)] <- as.character(col_tcga$tcgaid)
    115.   rnaExpr <<- mRNA_df
    117.   save(rnaExpr,file = "./TCGA_miRNAexpr.Rdata")
    119.   print(paste("your RNA exprset has been saved at: ", getwd(), sep = ""))

    获取更多R语言知识，请关注公众号：医学和生信笔记

    医学和生信笔记 公众号主要分享：1.医学小知识、肛肠科小知识；2.R语言和Python相关的数据分析、可视化、机器学习等；3.生物信息学学习资料和自己的学习笔记！