Entering edit mode
5.4 years ago
Freek
▴
60
Hi,
I have used RSEM to estimate Transcript abundances in a sample. RSEM uses STAR to do the mapping (if asked) and optionally can output a BAM file where the reads are aligned against the genome and not the transcriptome (this genome file ends in .STAR.genome.bam and is generated with the option "--star-output-genome-bam "). Now I was wondering, I want to use RSeQC to look at various quality related aspects of my BAM file, but which one should I use as input? The one aligned against the transcriptome (.transcript.bam) of the one aligned against the genome (.STAR.genome.bam)?
Highest regards,
Freek.
Did you ever find the answer to this question? I have been used genomic BAM files when processing data through RSeQC, but I wanted to double check whether this was the correct approach.
Hi, thanx for coming back at this. 16 months later, 16 months wiser :)
I use the genome BAM file, although I never tested the transcriptome BAM file, I'm assuming, since the transcriptome BAM was created with the transcriptome as a reference, that there are no reads mapped to anything other than exons (valid transcripts) in said BAM file, this would mean that RSeQC can not determine the number of intronic reads or reads on intergenic regions etc. Since RSeQC uses it's own gene/transcript definitions (I downloaded the BED file they offer), I think it only makes sense to apply RSeQC to a genome BAM.
Still, I might just feed RSeQC the transciptome BAM once and see what happens, in the days to come :)
Hi @benpkeith and @Freek Just stumbled upon this question and I also have the same issue. RSeQC - does it make sense to run it on the genome BAM when alignment was to transcriptome with STAR? --- did you ever figure out the correct way of running RSeQC (genome BAM versus transcriptome BAM)?
I stand by my earlier answer: Use the genome Bam. Did you ever feed it the Transcriptome bam? I think that at least several parameters (ie related to intronic and intergenic reads) will be non-sensical.