Overview

This package leverages the MultiAssayExperiment and curatedTCGAData packages to download the relevant data, and assemble a singular, pan-cancer, FacileTcgaDataSet using a subset of the assays made availalbe from those resources.

Gene-level quantitaion (expression, CNV, etc) will be mapped to the ensembl gene-level universe. Entries that cannot be mapped that way will be dropped.

Datasets

We will create a dataset across all indications using the following assays:

mRNA Abundance

The following assays can be pulled from the curatedTCGAData package:

• RNASeq2GeneNorm: fpkm values (sad)
• RNASeqGene: v2 fpkm values?
• miRNASeqGene: miRNA-seq data

We might consider getting gene and transcript level counts from the recount2 project, though:

https://jhubiostatistics.shinyapps.io/recount/

They used Gencode v25 GFF3 annotations, which we can parse with the utility funcionts in GemomicsStudyDb

Copy Number

• GISTIC_AllByGene: real valued copy number per gene
• GISTIC_ThresholdedByGene: threshold (duplication / deletion / normal) gene-level CNV scores

GenomicVariants

• Mutation: mutation status? What does this look like?

Other

• RPPAArray: Reverse Phase Protein Array (NOISY)
• Methylation, Methylation_methyl27, or Methylation_methyl450

Individual indication will be downloaded separately and assembled into one final FacileTcgaDataSet.

Data Assembly

Let’s follow along with the curatedTCGAData vignette to figure out how to do this.