RNA Seq analysis

[richaM]

I want to extract differential expressed genes using LIMMA from RNA seq data for three cancer types viz breast, lung and prostate. These data should have tumor and normal samples. I have read some papers which have used data from TCGA. BUt now TCGA has linked to Genomics Data Commons and all data are not open access. All BAM files are under controlled access. Also, LIMMA requires raw read counts for analysis. I an new to RNA seq data and analysis. Can anyone help, where should I get these data and what format should it be, as I have read that TPM, FPKM normalized values cannot be input to LIMMA.

[GenoMax]

For reference cross-posted: https://www.biostars.org/p/250946/

[aprice67]

I'm not sure what you'll have to do to get your data, but I can give you some advice on the RNA-seq part.

If you're starting with BAM files that means your reads are already aligned, if you start with reads (.fastq) you can map them yourself using any splice aware mapping tool. Assuming you have bam files though, you'll want to use a tool that extracts raw counts from the bam files. You'll need an annotation file (.gtf or gff format probably) and a counting tool. htseq-counts or featureCounts (in the RSubread package) are both good and widely used tools for extracting raw counts from BAM files, so start there and extract your counts.

Once you have gotten that far, then there are alot of tools like limma, deseq2, or edgeR that can use that data for analysis.