GDCRNATools – An R package for downloading, organizing, and integrative analyzing lncRNA, mRNA, and miRNA data in GDC

Introduction

The Genomic Data Commons (GDC) maintains standardized genomic, clinical, and biospecimen data from National Cancer Institute (NCI) programs including The Cancer Genome Atlas (TCGA) and Therapeutically Applicable Research To Generate Effective Treatments (TARGET), It also accepts high quality datasets from non-NCI supported cancer research programs, such as genomic data from the Foundation Medicine.

GDCRNATools is an R package which provides a standard, easy-to-use and comprehensive pipeline for downloading, organizing, and integrative analyzing RNA expression data in the GDC portal with an emphasis on deciphering the lncRNA-mRNA related ceRNA regulatory network in cancer.

Many analyses can be perfomed using GDCRNATools, including differential gene expression analysis (limma(Ritchie et al. 2015), edgeR(Robinson, McCarthy, and Smyth 2010), and DESeq2(Love, Huber, and Anders 2014)), univariate survival analysis (CoxPH and KM), competing endogenous RNA network analysis (hypergeometric test, Pearson correlation analysis, regulation similarity analysis, sensitivity Pearson partial correlation), and functional enrichment analysis(GO, KEGG, DO). Besides some routine visualization methods such as volcano plot, scatter plot, and bubble plot, etc., three simple shiny apps are developed in GDCRNATools allowing users visualize the results on a local webpage.

This user-friendly package allows researchers perform the analysis by simply running a few functions and integrate their own pipelines such as molecular subtype classification, weighted correlation network analysis (WGCNA)(Langfelder and Horvath 2008), and TF-miRNA co-regulatory network analysis, etc. into the workflow easily.

Installation

GDCRNATools is now under review in Bioconductor. Users can install the package locally.

On Windows system

Download the package GDCRNATools_0.99.0.tar.gz
Make sure you have Rtools installed
ADD R and Rtools to the Path Variable on the Environment Variables panel, including

c:\program files\Rtools\bin

c:\program files\Rtools\gcc-4.6.3\bin

c:\program files\R\R.3.x.x\bin\i386

c:\program files\R\R.3.x.x\bin\x64
Open a command prompt. Type R CMD INSTALL GDCRNATools_0.99.0.tar.gz

On Linux and Mac systems

Run the following command in R

install.packages('GDCRNATools_0.99.0.tar.gz', repos = NULL, type='source')

If GDCRNATools cannot be installed due to the lack of dependencies, please run the following code ahead to install those pacakges simutaneously or separately:

source("https://bioconductor.org/biocLite.R")

### install packages simutaneously ###
biocLite(c('limma', 'edgeR', 'DESeq2', 'clusterProfiler', 'DOSE', 'org.Hs.eg.db', 'biomaRt', 'BiocParallel'))
install.packages(c('shiny', 'jsonlite', 'rjson', 'survival', 'survminer', 'ggplot2', 'gplots', 'Hmisc'))

### install packages seperately ###
biocLite('limma')
biocLite('edgeR')
biocLite('DESeq2')
biocLite('clusterProfiler')
biocLite('DOSE')
biocLite('org.Hs.eg.db')
biocLite('biomaRt')
biocLite('BiocParallel')

install.packages('shiny')
install.packages('jsonlite')
install.packages('rjson')
install.packages('survival')
install.packages('survminer')
install.packages('ggplot2')
install.packages('gplots')
install.packages('Hmisc')

Manual

A simply manual of GDCRNATools is available below. Users are also highly recommended to download the comprhensive manual in .html format and view on local computer GDCRNATools Manual

1 Data download

Two methods are provided for downloading Gene Expression Quantification (HTSeq-Counts), Isoform Expression Quantification (BCGSC miRNA Profiling), and Clinical (Clinical Supplement) data:

Manual download Step1: Download GDC Data Transfer Tool on the GDC website Step2: Add data to the GDC cart, then download manifest file and metadata of the cart Step3: Download data using gdcRNADownload() function by providing the manifest file
Automatic download Download GDC Data Transfer Tool, manifest file, and data automatically by specifying the project.id and data.type in gdcRNADownload() function for RNAseq and miRNAs data, and in gdcClinicalDownload() function for clinical data

Users can also download data from GDC using the API method developed in TCGAbiolinks(Colaprico et al. 2016) or using TCGA-Assembler(Zhu, Qiu, and Ji 2014)

1.1 Manual download

1.1.1 Installation of GDC Data Transfer Tool gdc-client

Download GDC Data Transfer Tool from the GDC website and unzip the file

1.1.2 Download manifest file and metadata from GDC Data Portal

1.1.3 Download data

####### Download RNAseq data #######



gdcRNADownload(manifest  = 'TCGA-PRAD/TCGA-PRAD.RNAseq.gdc_manifest.2017-11-23T14-40-52.txt',



               directory = 'TCGA-PRAD/RNAseq')



 



####### Download miRNAs data #######



gdcRNADownload(manifest  = 'TCGA-PRAD/TCGA-PRAD.miRNAs.gdc_manifest.2017-11-22T15-36-57.txt',



               directory = 'TCGA-PRAD/miRNAs')



 



####### Download Clinical data #######



gdcRNADownload(manifest  = 'TCGA-PRAD/TCGA-PRAD.Clinical.gdc_manifest.2017-11-23T14-42-01.txt',



               directory = 'TCGA-PRAD/Clinical')

1.2 Automatic download

gdcRNADownload() will download HTSeq-Counts data if data.type='RNAseq' and download BCGSC miRNA Profiling data if data.type='miRNAs'. project.id argument is required to be provided.
gdcClinicalDownload() download clinical data in .xml format automatically by simply specifying the project.idargument.

1.2.1 Download RNAseq/miRNAs data

####### Download RNAseq data #######



gdcRNADownload(project.id     = 'TCGA-PRAD', 



               data.type      = 'RNAseq', 



               write.manifest = FALSE,



               directory      = 'TCGA-PRAD/RNAseq')



 



####### Download miRNAs data #######



gdcRNADownload(project.id     = 'TCGA-PRAD', 



               data.type      = 'miRNAs', 



               write.manifest = FALSE,



               directory      = 'TCGA-PRAD/miRNAs')

1.2.2 Download clinical data

####### Download clinical data #######



gdcClinicalDownload(project.id     = 'TCGA-PRAD',  



                    write.manifest = FALSE,



                    directory      = 'TCGA-PRAD/Clinical')

2 Data organization

2.1 Parse metadata

Metadata can be parsed by either providing the metadata file that is downloaded in the data download step, or specifying the project.id and data.type in gdcParseMetadata() function to obtain information of data in the manifest file to facilitate data organization and basic clinical information of patients such as age, stage and gender, etc. for data analysis.

2.1.1 Parse metadata by providing the metadata file

####### Parse RNAseq metadata #######



metaMatrix.RNA <- gdcParseMetadata(metafile='TCGA-PRAD/TCGA-PRAD.RNAseq.metadata.2017-11-23T17-23-59.json')



 



####### Parse miRNAs metadata #######



metaMatrix.MIR <- gdcParseMetadata(metafile='TCGA-PRAD/TCGA-PRAD.miRNAs.metadata.2017-11-23T17-33-55.json')

2.1.2 Parse metadata by specifying project.id and data.type

####### Parse RNAseq metadata #######



metaMatrix.RNA <- gdcParseMetadata(project.id = 'TCGA-PRAD',



                                   data.type  = 'RNAseq', 



                                   write.meta = FALSE)
####### Parse miRNAs metadata #######



metaMatrix.MIR <- gdcParseMetadata(project.id = 'TCGA-PRAD',



                                   data.type  = 'miRNAs', 



                                   write.meta = FALSE)

2.2 Filter samples

2.2.1 Filter duplicated samples

Only one sample would be kept if the sample had been sequenced more than once by gdcFilterDuplicate().

####### Filter duplicated samples in RNAseq metadata #######



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



metaMatrix.MIR <- gdcFilterDuplicate(metaMatrix.MIR)

2.2.2 Filter non-Primary Tumor and non-Solid Tissue Normal samples

Samples that are neither Primary Tumor (code: 01) nor Solid Tissue Normal (code: 11) would be filtered out by gdcFilterSampleType().

####### Filter non-Primary Tumor and non-Solid Tissue Normal samples in RNAseq metadata #######



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



metaMatrix.MIR <- gdcFilterSampleType(metaMatrix.MIR)

2.3 Merge data

gdcRNAMerge() merges raw counts data of RNAseq to a single expression matrix with rows are Ensembl id and columns are samples. Total read counts for 5p and 3p strands of miRNAs can be processed from isoform quantification files and then merged to a single expression matrix with rows are miRBase v21 identifiers and columns are samples.
gdcClinicalMerge() merges clinical data to a dataframe with rows are patient id and columns are clinical traits. If key.info=TRUE, only those most commonly used clinical traits will be reported, otherwise, all the clinical information will be reported.

2.3.1 Merge RNAseq/miRNAs data

####### Merge RNAseq data #######



rnaMatrix <- gdcRNAMerge(metadata  = metaMatrix.RNA, 



                         path      = 'TCGA-PRAD/RNAseq/', 



                         data.type = 'RNAseq')



 



####### Merge miRNAs data #######



mirMatrix <- gdcRNAMerge(metadata  = metaMatrix.MIR,



                         path      = 'TCGA-PRAD/miRNAs/',



                         data.type = 'miRNAs')

2.3.2 Merge clinical data

####### Merge clinical data #######



clinicalDa <- gdcClinicalMerge(path = 'TCGA-PRAD/Clinical/', key.info = TRUE)

2.4 TMM normalization and voom transformation

It has repeatedly shown that normalization is a critical way to ensure accurate estimation and detection of differential expression (DE) by removing systematic technical effects that occur in the data(Robinson and Oshlack 2010). TMM normalization is a simple and effective method for estimating relative RNA production levels from RNA-seq data. Voom is moreover faster and more convenient than existing RNA-seq methods, and converts RNA-seq data into a form that can be analyzed using similar tools as for microarrays(Law et al. 2014).

By running gdcVoomNormalization() function, raw counts data would be normalized by TMM method implemented in edgeR(Robinson, McCarthy, and Smyth 2010) and further transformed by the voom method provided in limma(Ritchie et al. 2015). Low expression genes (logcpm < 1 in more than half of the samples) will be filtered out by default. All the genes can be kept by setting filter=TRUE in the gdcVoomNormalization().

####### RNAseq data #######



rnaExpr <- gdcVoomNormalization(counts = rnaMatrix, filter = FALSE)



 



####### miRNAs data #######



mirExpr <- gdcVoomNormalization(counts = mirMatrix, filter = FALSE)

3. Differential gene expression analysis

gdcDEAnalysis(), a convenience wrapper, provides three widely used methods limma(Ritchie et al. 2015), edgeR(Robinson, McCarthy, and Smyth 2010), and DESeq2(Love, Huber, and Anders 2014) to identify differentially expressed genes (DEGs) or miRNAs between any two groups defined by users. Note that DESeq2(Love, Huber, and Anders 2014) maybe slow with a single core. Multiple cores can be specified with the nCore argument if DESeq2(Love, Huber, and Anders 2014) is in use. Users are encouraged to consult the vignette of each method for more detailed information.

3.1 DE analysis

DEGAll <- gdcDEAnalysis(counts     = rnaMatrix, 



                        group      = metaMatrix.RNA$sample_type, 



                        comparison = 'PrimaryTumor-SolidTissueNormal', 



                        method     = 'limma')

3.2 Report DE genes/miRNAs

All DEGs, DE long non-coding genes, DE protein coding genes and DE miRNAs could be reported separately by setting geneType argument in gdcDEReport(). Gene symbols and biotypes based on the Ensembl 90 annotation are reported in the output.

### 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')

3.3 DEG visualization

Volcano plot and bar plot are used to visualize DE analysis results in different manners by gdcVolcanoPlot() and gdcBarPlot() functions, respectively . Hierarchical clustering on the expression matrix of DEGs can be analyzed and plotted by the gdcHeatmap() function.

3.3.1 Volcano plot

gdcVolcanoPlot(DEGAll)

3.3.2 Barplot

gdcBarPlot(deg = deALL, angle = 45, data.type = 'RNAseq')

3.3.3 Heatmap

Heatmap is generated based on the heatmap.2() function in gplots package.

degName = rownames(deALL)
gdcHeatmap(deg.id = degName, metadata = metaMatrix.RNA, rna.expr = rnaExpr)

4 Competing endogenous RNAs network analysis

Three criteria are used to determine the competing endogenous interactions between lncRNA-mRNA pairs:

The lncRNA and mRNA must share significant number of miRNAs
Expression of lncRNA and mRNA must be positively correlated
Those common miRNAs should play similar roles in regulating the expression of lncRNA and mRNA

4.1 Hypergeometric test

Hypergenometric test is performed to test whether a lncRNA and mRNA share many miRNAs significantly.

A newly developed algorithm spongeScan(Furi’o-Tar’i et al. 2016) is used to predict MREs in lncRNAs acting as ceRNAs. Databases such as starBase v2.0(J.-H. Li et al. 2014), miRcode(Jeggari, Marks, and Larsson 2012) and mirTarBase release 7.0(Chou et al. 2017) are used to collect predicted and experimentally validated miRNA-mRNA and/or miRNA-lncRNA interactions. Gene IDs in these databases are updated to the latest Ensembl 90 annotation of human genome and miRNAs names are updated to the new release miRBase 21 identifiers. Users can also provide their own datasets of miRNA-lncRNA and miRNA-mRNA interactions.

The figure and equation below illustrate how the hypergeometric test works

$p=1-\sum_{k=0}^m \frac{\binom{K}{k}\binom{N-K}{n-k}}{\binom{N}{n}}$

here is the number of shared miRNAs, is the total number of miRNAs in the database, is the number of miRNAs targeting the lncRNA, is the number of miRNAs targeting the protein coding gene.

4.2 Pearson correlation analysis

Pearson correlation coefficient is a measure of the strength of a linear association between two variables. As we all know, miRNAs are negative regulators of gene expression. If more common miRNAs are occupied by a lncRNA, less of them will bind to the target mRNA, thus increasing the expression level of mRNA. So expression of the lncRNA and mRNA in a ceRNA pair should be positively correlated.

4.3 Regulation pattern analysis

Two methods are used to measure the regulatory role of miRNAs on the lncRNA and mRNA:

Regulation similarity

We defined a measurement regulation similarity score to check the similarity between miRNAs-lncRNA expression correlation and miRNAs-mRNA expression correlation.

![$$Regulation\ similarity\ score = 1-\frac{1}{M} \sum_{k=1}^M [{\frac{|corr(m_k,l)-corr(m_k,g)|}{|corr(m_k,l)|+|corr(m_k,g)|}}]^M$$](https://camo.githubusercontent.com/f265c1e6b65a0e22912fa0580af4c361dae54567/68747470733a2f2f6c617465782e636f6465636f67732e636f6d2f6769662e6c617465783f2424526567756c6174696f6e5c2673706163653b73696d696c61726974795c2673706163653b73636f72652673706163653b3d2673706163653b312d5c667261637b317d7b4d7d2673706163653b5c73756d5f7b6b3d317d5e4d2673706163653b5b7b5c667261637b7c636f7272286d5f6b2c6c292d636f7272286d5f6b2c67297c7d7b7c636f7272286d5f6b2c6c297c2b7c636f7272286d5f6b2c67297c7d7d5d5e4d2424)

where is the total number of shared miRNAs, is the th shared miRNAs, and represents the Pearson correlation between the th miRNA and lncRNA, the th miRNA and mRNA, respectively

Sensitivity correlation

Sensitivity correlation is defined by Paci et al.(2014) to measure if the correlation between a lncRNA and mRNA is mediated by a miRNA in the lncRNA-miRNA-mRNA triplet. We take average of all triplets of a lncRNA-mRNA pair and their shared miRNAs as the sensitivity correlation between a selected lncRNA and mRNA.

$Sensitivity\ correlation = corr(l,g)-\frac{1}{M}\sum_{k=1}^M {\frac{corr(l,g)-corr(m_k,l)corr(m_k,g)}{\sqrt{1-corr(m_k,l)^2}\sqrt{1-corr(m_k,g)^2}}}$

where is the total number of shared miRNAs, is the th shared miRNAs, , and represents the Pearson correlation between the long non-coding RNA and the protein coding gene, the kth miRNA and lncRNA, the kth miRNA and mRNA, respectively

The hypergeometric test of shared miRNAs, expression correlation analysis of lncRNA-mRNA pair, and regulation pattern analysis of shared miRNAs are all implemented in the gdcCEAnalysis() function.

ceOutput <- gdcCEAnalysis(lnc         = rownames(deLNC), 



                          pc          = rownames(dePC), 



                          lnc.targets = 'starBase', 



                          pc.targets  = 'starBase', 



                          rna.expr    = rnaExpr, 



                          mir.expr    = mirExpr)

4.4 ceRNAs visualization

4.4.1 Correlation plot

gdcCorPlot(gene1    = 'ENSG00000234456', 



           gene2    = 'ENSG00000105971',



           rna.expr = rnaExpr,



           metadata = metaMatrix.RNA)

4.4.2 Correlation plot on a local webpage by shinyCorplot

Typing and running gdcCorPlot() for each pair of lncRNA-mRNA is bothering when multiple pairs are being interested in. shinyCorPlot() , a interactive plot function based on shiny package, can be easily operated by just clicking the genes in each drop down box (in the GUI window). By running shinyCorPlot() function, a local webpage would pop up and correlation plot between a lncRNA and mRNA would be automatically shown.

shinyCorPlot(gene1    = rownames(deLNC), 



             gene2    = rownames(dePC), 



             rna.expr = rnaExpr, 



             metadata = metaMatrix.RNA)

4.4.3 Network visulization in Cytoscape

lncRNA-miRNA-mRNA interactions can be reported by the gdcExportNetwork() and visualized 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')

5 Univariate survival analysis

Two methods are provided to perform univariate survival analysis: Cox Proportional-Hazards (CoxPH) model and Kaplan Meier (KM) analysis based on the survival package. CoxPH model considers expression value as continous variable while KM analysis divides patients into high-expreesion and low-expression groups by a user-defined threshold such as median or mean. gdcSurvivalAnalysis() take a list of genes as input and report the hazard ratio, 95% confidence intervals, and test significance of each gene on overall survival.

5.1 CoxPH analysis

####### CoxPH analysis #######



survOutput <- gdcSurvivalAnalysis(gene     = rownames(deALL), 



                                  method   = 'coxph', 



                                  rna.expr = rnaExpr, 



                                  metadata = metaMatrix.RNA)

5.2 KM analysis

####### KM analysis #######



survOutput <- gdcSurvivalAnalysis(gene     = rownames(deALL), 



                                  method   = 'KM', 



                                  rna.expr = rnaExpr, 



                                  metadata = metaMatrix.RNA, 



                                  sep      = 'median')

5.3 KM analysis visualization

5.5.1 KM plot

KM survival curves are ploted using the gdcKMPlot() function which is based on the R package survminer.

gdcKMPlot(gene     = 'ENSG00000197275', 



          rna.expr = rnaExpr, 



          metadata = metaMatrix.RNA, 



          sep      = 'median')

5.3.2 KM plot on a local webpage by shinyKMPlot

The shinyKMPlot() function is also a simply shiny app which allow users view KM plots of all genes of interests on a local webpackage conveniently.

shinyKMPlot(gene = rownames(deALL), rna.expr = rnaExpr, metadata = metaMatrix.RNA)

6 Functional enrichment analysis

One of the main uses of the GO is to perform enrichment analysis on gene sets. For example, given a set of genes that are up-regulated under certain conditions, an enrichment analysis will find which GO terms are over-represented (or under-represented) using annotations for that gene set and pathway enrichment can also be applied afterwards.

6.1 GO, KEGG and DO analyses

gdcEnrichAnalysis() can perform Gene ontology (GO), Kyoto Encyclopedia of Genes and Genomes (KEGG) and Disease Ontology (DO) functional enrichment analyses of a list of genes simultaneously. GO and KEGG analyses are based on the R/Bioconductor packages clusterProfilier(Yu et al. 2012) and DOSE(Yu et al. 2015). Redundant GO terms can be removed by specifying simplify=TRUE in the gdcEnrichAnalysis() function which uses the simplify() function in the clusterProfilier(Yu et al. 2012) package.

enrichOutput <- gdcEnrichAnalysis(gene = rownames(deALL), simplify = TRUE)

6.2 Enrichment visualization

The output generated by gdcEnrichAnalysis() can be used for visualization in the gdcEnrichPlot() function by specifying type,category and numTerms arguments.

6.2.1 GO barplot

gdcEnrichPlot(enrichOutput, type = 'bar', category = 'GO', num.terms = 10)

6.2.2 GO bubble plot

gdcEnrichPlot(enrichOutput, type='bubble', category='GO', num.terms = 10)

6.2.3 KEGG/DO barplot

gdcEnrichPlot(enrichment = enrichOutput, 



              type       = 'bar', 



              category   = 'KEGG', 



              bar.color  = 'chocolate1', 



              num.terms  = 20)

gdcEnrichPlot(enrichment = enrichOutput, 



              type       = 'bar', 



              category   = 'DO', 



              bar.color  = 'dodgerblue', 



              num.terms  = 20)

6.2.4 KEGG/DO bubble plot

gdcEnrichPlot(enrichOutput, category='KEGG',type = 'bubble', num.terms = 20)

gdcEnrichPlot(enrichOutput, category='DO',type = 'bubble', num.terms = 20)

6.2.5 Pathview

Users can visualize a pathway map with pathview() function in the pathview(Luo and Brouwer 2013) package. It displays related many-genes-to-many-terms on 2-D view, shows by genes on BioCarta & KEGG pathway maps. Gradient colors can be used to indicate if genes are up-regulated or down-regulated.

deg <- deALL$logFC



names(deg) <- rownames(deALL)
library(pathview)



hsa04022 <- pathview(gene.data   = deg,



                     pathway.id  = "hsa04022",



                     species     = "hsa",



                     gene.idtype = 'ENSEMBL',



                     limit       = list(gene=max(abs(geneList)), cpd=1))

6.2.6 View pathway maps on a local webpage by shinyPathview

shinyPathview() allows users view and download pathways of interests by simply selecting the pathway terms on a local webpage.

pathways <- as.character(enrichOutput$Terms[enrichOutput$Category=='KEGG'])
shinyPathview(deg, pathways = pathways, directory = 'pathview')

请关注“恒诺新知”微信公众号，感谢“R语言“，”数据那些事儿“，”老俊俊的生信笔记“，”冷🈚️思“，“珞珈R”，“生信星球”的支持！