salmon - tximport - deseq2 gene level workflow

Question

[jomo018]

I am applying salmon quant with --geneMap option to quantify RNA-SEQ to gene level count files. I plan to proceed with DE using Deseq2.

Is it OK to use txIn and txOut in tximport as follows:

tximport(files, type="salmon", countsFromAbundance="no", txIn=T , txOut=T ).

In other words, apply the gene count files as if these are transcripts and avoid any upper level aggregation. I did not find the natural way of applying gene counts to tximport.

[ATpoint]

Why don't you make it as simple as possible and run salmon with the default parameters on your fastq files, followed by standard tximport to aggregate tx to gene level?

[jomo018]

I believe there is contamination on one chromosome (M) which I want to get rid off after alignment but before DE. It's easier to do on the gene level right after Salmon. There are other ways to do that, but I thought tximport on gene level would be straightforward.