In LiNk-NY/curatedTCGAManu: Presents example analyses using curatedTCGAData, MultiAssayExperiment, and TCGAutils

knitr::opts_chunk$set(
  collapse = TRUE,
  comment = "#>",
  cache = TRUE,
  out.width = "100%"
)

curatedTCGAManu

Multi-omic integration of public oncology databases in Bioconductor

This repository contains scripts and datasets for the curatedTCGAData + cBioPortalData manuscript.

Overview

We provide several examples to demonstrate the powerful and flexible analysis environment provided. These analyses, previously only achievable through a significant investment of time and bioinformatic training, become straight- forward analysis exercises provided as vignettes.

Key Points

• Key objective: To provide flexible, integrated multi-omic representations of public oncology databases in R/Bioconductor with grealy reduced data management overhead.

• Knowledge generated: Our Bioconductor software packages provide a novel approach to lower barriers to analysis and tool development for the TCGA and cBioPortal databases.

• Relevance: Our tools provide flexible, programmatic analysis of hundreds of fully integrated multi’omic oncology datasets within an ecosystem of multi-omic analysis tools.

Key Packages

• MultiAssayExperiment
• curatedTCGAData
• TCGAutils
• cBioPortalData
• EnrichmentBrowser
• GSEABenchmarkeR
• RTCGAToolbox

Installation

Until the release of Bioconductor 3.11 (scheduled for April 28, 2020), it is strongly recommended to use the devel version of Bioconductor. That version can be installed the traditional way or by using the Docker container.

Additionally until the release of Bioconductor 3.11, cBioPortalData must be installed from GitHub as shown in the following code chunk which installs all necessary packages either directly or as dependencies. Note that this code chunk is not evaluated, because installation only needs to be performed once.

BiocManager::install(c("cBioPortalData", "LiNk-NY/curatedTCGAManu"))

Vignette Build

Because of the size of the data, it is recommended that the vignettes be built individually. If using RStudio, the user can simply open the vignette and press the knit button. Otherwise, the package can be built completely with vignettes by doing:

    R CMD build curatedTCGAManu

in the command line or

BiocManager::install("Link-NY/curatedTCGAManu", build_vignettes = TRUE)

in R.

Loading packages

library(curatedTCGAData)
library(cBioPortalData)
library(TCGAutils)
library(GenomicDataCommons)
library(rtracklayer)

Note. For clarity, we include library commands within the supplemental code chunks.

Figures

Ease-of-use schematic

Figure 1

Pan-Cancer OncoPrint Plot

Figure 2

Differential Expression and GSEA PanCan

Figure 3

Figure 4

Example Multi-omic Analyses

Figure 5

Figure 6

Supplement Reference

Figure S1

Data provenance and package network

Figure S2A

Example code for installing and downloading TCGA data using curatedTCGAData

if (!requireNamespace("BiocManager", quietly = TRUE))
    install.packages("BiocManager")

if (!requireNamespace("curatedTCGAData", quietly = TRUE))
    BiocManager::install("curatedTCGAData")

## Glioblastoma Multiforme (GBM)
library(curatedTCGAData)
curatedTCGAData(diseaseCode = "GBM", assays = "RNA*", dry.run = FALSE)

Figure S2B

Example cBioPortalData code for downloading and exporting TCGA data from cBioPortal.org and via the cBioPortal API

## installation
if (!requireNamespace("cBioPortalData", quietly = TRUE))
    BiocManager::install("cBioPortalData")

library(cBioPortalData)

## https://cbioportal.org/datasets (Bulk data method)
gbm <- cBioDataPack("gbm_tcga")

## https://cBioPortal.org/api (API method)
cBio <- cBioPortal()

## use exportClass() with the result to save data files
tcga_gbm <- cBioPortalData(cBio, studyId = "gbm_tcga", genePanelId = "IMPACT341")

tcga_gbm

exportClass(tcga_gbm, dir = tempdir(), fmt = "csv")

Figure S2C

Example hg19 to hg38 liftover procedure using Bioconductor tools

liftchain <- "http://hgdownload.cse.ucsc.edu/goldenpath/hg19/liftOver/hg19ToHg38.over.chain.gz"
cloc38 <- file.path(tempdir(), gsub("\\.gz", "", basename(liftchain)))
dfile <- tempfile(fileext = ".gz")
download.file(liftchain, dfile)
R.utils::gunzip(dfile, destname = cloc38, remove = FALSE)

library(rtracklayer)
chain38 <- suppressMessages( import.chain(cloc38) )

## Run bulk data download (from S2B) to create gbm object
if (!exists("gbm")) gbm <- cBioPortalData::cBioDataPack("gbm_tcga")

mutations <- gbm[["mutations_extended"]]
seqlevelsStyle(mutations) <- "UCSC"

ranges38 <- liftOver(rowRanges(mutations), chain38)

Figure S3

Example code for downloading data via GenomicDataCommons and loading with TCGAutils

library(TCGAutils)
library(GenomicDataCommons)

## GenomicDataCommons
query <- files(legacy = TRUE) %>%
    filter( ~ cases.project.project_id == "TCGA-COAD" &
        data_category == "Gene expression" &
        data_type == "Exon quantification" )

fileids <- manifest(query)$id[1:4]
exonfiles <- gdcdata(fileids, use_cached = FALSE)

## TCGAutils
makeGRangesListFromExonFiles(exonfiles, nrows = 4)

Figure S4

Correlated principal components across experimental assays in adrenocortical carcinoma (ACC)

LiNk-NY/curatedTCGAManu documentation built on July 22, 2020, 4:06 p.m.