Case Study

library(GDCRNATools)

Download RNA and mature miRNA expression data

project <- 'TCGA-CHOL'
rnadir <- paste(project, 'RNAseq', sep='/')
mirdir <- paste(project, 'miRNAs', sep='/')

####### Download RNAseq data #######
gdcRNADownload(project.id     = 'TCGA-CHOL', 
               data.type      = 'RNAseq', 
               write.manifest = FALSE,
               method         = 'gdc-client',
               directory      = rnadir)

####### Download mature miRNA data #######
gdcRNADownload(project.id     = 'TCGA-CHOL', 
               data.type      = 'miRNAs', 
               write.manifest = FALSE,
               method         = 'gdc-client',
               directory      = mirdir)

Download clinical data

####### Download clinical data #######
clinicaldir <- paste(project, 'Clinical', sep='/')
gdcClinicalDownload(project.id     = 'TCGA-CHOL', 
                    write.manifest = FALSE,
                    method         = 'gdc-client',
                    directory      = clinicaldir)

Data organization and DE analysis

Parse metadata

####### Parse RNAseq metadata #######
metaMatrix.RNA <- gdcParseMetadata(project.id = 'TCGA-CHOL',
                                   data.type  = 'RNAseq', 
                                   write.meta = FALSE)

####### Filter duplicated samples in RNAseq metadata #######
metaMatrix.RNA <- gdcFilterDuplicate(metaMatrix.RNA)

####### Filter non-Primary Tumor and non-Solid Tissue Normal samples in RNAseq metadata #######
metaMatrix.RNA <- gdcFilterSampleType(metaMatrix.RNA)

####### Parse miRNAs metadata #######
metaMatrix.MIR <- gdcParseMetadata(project.id = 'TCGA-CHOL',
                                   data.type  = 'miRNAs', 
                                   write.meta = FALSE)

####### Filter duplicated samples in miRNAs metadata #######
metaMatrix.MIR <- gdcFilterDuplicate(metaMatrix.MIR)

####### Filter non-Primary Tumor and non-Solid Tissue Normal samples in miRNAs metadata #######
metaMatrix.MIR <- gdcFilterSampleType(metaMatrix.MIR)

Merge raw counts data

####### Merge RNAseq data #######
rnaCounts <- gdcRNAMerge(metadata  = metaMatrix.RNA, 
                         path      = rnadir, # the folder in which the data stored
                         organized = FALSE, # if the data are in separate folders
                         data.type = 'RNAseq')

####### Merge miRNAs data #######
mirCounts <- gdcRNAMerge(metadata  = metaMatrix.MIR,
                         path      = mirdir, # the folder in which the data stored
                         organized = FALSE, # if the data are in separate folders
                         data.type = 'miRNAs')

Merge clinical data

####### Merge clinical data #######
clinicalDa <- gdcClinicalMerge(path = clinicaldir, key.info = TRUE)
clinicalDa[1:6,5:10]

TMM normalization and voom transformation

####### Normalization of RNAseq data #######
rnaExpr <- gdcVoomNormalization(counts = rnaCounts, filter = FALSE)

####### Normalization of miRNAs data #######
mirExpr <- gdcVoomNormalization(counts = mirCounts, filter = FALSE)

Differential gene expression analysis

DEGAll <- gdcDEAnalysis(counts     = rnaCounts, 
                        group      = metaMatrix.RNA$sample_type, 
                        comparison = 'PrimaryTumor-SolidTissueNormal', 
                        method     = 'limma')

### All DEGs
deALL <- gdcDEReport(deg = DEGAll, gene.type = 'all')

### DE long-noncoding
deLNC <- gdcDEReport(deg = DEGAll, gene.type = 'long_non_coding')

### DE protein coding genes
dePC <- gdcDEReport(deg = DEGAll, gene.type = 'protein_coding')

Competing endogenous RNAs network analysis

ceOutput <- gdcCEAnalysis(lnc         = rownames(deLNC), 
                          pc          = rownames(dePC), 
                          lnc.targets = 'starBase', 
                          pc.targets  = 'starBase', 
                          rna.expr    = rnaExpr, 
                          mir.expr    = mirExpr)

Network visulization in Cytoscape

ceOutput2 <- ceOutput[ceOutput$hyperPValue<0.01 & 
    ceOutput$corPValue<0.01 & ceOutput$regSim != 0,]

edges <- gdcExportNetwork(ceNetwork = ceOutput2, net = 'edges')
nodes <- gdcExportNetwork(ceNetwork = ceOutput2, net = 'nodes')

write.table(edges, file='edges.txt', sep='\t', quote=F)
write.table(nodes, file='nodes.txt', sep='\t', quote=F)