Processing of RNA-Seq data

SRA files were converted into fastq format using fastq-dump from the SRA Toolkit (https://trace.ncbi.nlm.nih.gov/Traces/sra/sra.cgi? view=software). The reads from each library were aligned uniquely and independently with TopHat (version 2.0.6) (26) to the hg19 human reference genome which is available at http://bowtie-bio.sourceforge.net/bowtie2/index.shtml. All of the parameters used were default except for: –splice-mismatches 1, –mate-inner-dist 200, –library-type fr-unstranded and –transcriptome-index. Based on the mapped reads, FPKM values for transcripts were calculated using Cufflinks (version 2.0.2) (26) with default parameters except for –max-bundle-frags 500 000 and –multi-read-correct. Genes were discarded if they were not expressed at appreciable levels (FPKM > 1) in at least half of all the samples in the selected dataset.