TCGAbiolinks包下载更新后的TCGA数据

众所周知，今年TCGA数据库更新了一波，原来的HT-Counts现在变成了STAR-Counts。TCGABiolinks包的下载流程也发生了一些小小的变化。这里重新梳理一下TCGABiolinks的下载流程，供大家参考

一、加载R包

library(TCGAbiolinks)
library(SummarizedExperiment)

主要的R包主要是这么几个，其中SummarizedExperiment是为了提取不同类型（Counts/TPM……）的数据的。

二、下载数据

首先来查看一下TCGAbiolinks可以下载的数据类型

> getGDCprojects()$project_id
 [1] "EXCEPTIONAL_RESPONDERS-ER" "GENIE-GRCC"               
 [3] "GENIE-DFCI"                "GENIE-NKI"                
 [5] "GENIE-VICC"                "GENIE-UHN"                
 [7] "GENIE-MDA"                 "GENIE-MSK"                
 [9] "GENIE-JHU"                 "FM-AD"                    
[11] "OHSU-CNL"                  "MMRF-COMMPASS"            
[13] "ORGANOID-PANCREATIC"       "NCICCR-DLBCL"             
[15] "VAREPOP-APOLLO"            "CGCI-BLGSP"               
[17] "BEATAML1.0-CRENOLANIB"     "TRIO-CRU"                 
[19] "REBC-THYR"                 "TARGET-ALL-P2"            
[21] "TARGET-ALL-P1"             "CPTAC-2"                  
[23] "WCDT-MCRPC"                "CMI-ASC"                  
[25] "TCGA-READ"                 "TCGA-UCS"                 
[27] "CMI-MPC"                   "CMI-MBC"                  
[29] "BEATAML1.0-COHORT"         "TCGA-COAD"                
[31] "TCGA-CESC"                 "TCGA-PAAD"                
[33] "TCGA-ESCA"                 "TCGA-KIRP"                
[35] "TCGA-PCPG"                 "TCGA-HNSC"                
[37] "TCGA-BLCA"                 "TCGA-STAD"                
[39] "CTSP-DLBCL1"               "TCGA-SARC"                
[41] "TCGA-CHOL"                 "TCGA-LAML"                
[43] "TCGA-THYM"                 "TCGA-ACC"                 
[45] "TCGA-SKCM"                 "TCGA-LUAD"                
[47] "TCGA-LIHC"                 "TCGA-KIRC"                
[49] "TCGA-KICH"                 "TCGA-DLBC"                
[51] "TCGA-PRAD"                 "TCGA-OV"                  
[53] "TCGA-MESO"                 "TCGA-LUSC"                
[55] "TCGA-GBM"                  "TCGA-UVM"                 
[57] "TCGA-LGG"                  "HCMI-CMDC"                
[59] "TCGA-BRCA"                 "TARGET-RT"                
[61] "TARGET-CCSK"               "TCGA-TGCT"                
[63] "TARGET-NBL"                "CPTAC-3"                  
[65] "CGCI-HTMCP-CC"             "TARGET-ALL-P3"            
[67] "TARGET-OS"                 "TARGET-AML"               
[69] "TARGET-WT"                 "MP2PRT-WT"                
[71] "TCGA-THCA"                 "TCGA-UCEC"  

这里以结肠癌为例进行演示

COAD <- GDCquery(project = "TCGA-COAD",
         data.category = "Transcriptome Profiling",
         data.type = "Gene Expression Quantification",
         workflow.type = "STAR - Counts")
GDCdownload(COAD,method="api")

workflow.type这个参数，不管要下载的是TPM还是FPKM，都填STAR-Counts。不同类型的数据到后面再说。

经过漫长的等待数据终于下载下来了。文件默认存储在当前的工作目录下的GDCdata文件夹，当然也可以在GDCdownload函数里通过directory参数进行更改。

三、合并数据和提取数据

expr <- GDCprepare(query=COAD)

通过这条命令可以把上面下载到的数据整合成1个summarizedExperiment对象。
如果需要counts数据，可以直接从这个对象里提取

count <- as.data.frame(assay(expr))

如果需要counts格式以外的其他数据，则需要在这一步改一下参数

TPM <- as.data.frame(assay(expr,i = "tpm_unstrand"))

提取不同格式数据需要的参数在下面：

下载Counts i= "unstranded"
下载tpm i="tpm_unstrand"
下载fpkm i=" fpkm_unstrand"