dedup STAR transcriptome file using umi_tools

Question

[ATRX]

Hi,

I am interested in dedup the transcriptome output file (bulk RNA-seq) from STAR using umi_tools. I am using the following dedup function from umi_tools. Here is the command:

umi_tools dedup --paired
--stdin=B1-Cond1_Aligned.toTranscriptome.out.bam --log=B1-Cond1_dedup.txt --umi-separator=":" --output-stats=B1-Cond1_ > .dedup.bam

However, I am getting the following error:

ValueError: fetch called on bamfile without index

I don't think we can index transcriptome file from STAR. I read the umi_tools document and didn't much options for bulk RNA-seq libraries. Do you know what is the best way to dedup using umi_tools? Thanks!

[rfran010]

How about deduping based on the genomic bam, then filtering the transcriptome bam based on the reads leftover. Maybe some extra work, but might be faster if you can't figure out how to index/dedupe the transcriptome bam directly.

[ATRX]

Yes, you are right. I plan to do this. Thanks!

[ATRX]

Very helpful, thanks a lot!

Answer 1

[i.sudbery]

I don't know if any reason you shouldn't be able to index the bam produced by STAR, although you will need to sort them first.

[ATRX]

Thanks for the reply. STAR generates two different bam files. One is the genome based and the other one is the transcriptome based. I am able to dedup of genome based bam file but not for transcriptome based.

[i.sudbery]

You should definitely be able to use samtools ti sort and then index the transcriptome bam. We've done this many times.

[ATRX]

It worked. thanks a lot!

[GenoMax]

You can go ahead and accept @Ian answer to provide closure to this thread (green check mark).

[ATRX]

Thank you, I didn't know about the green check mark. I did it now, thanks!