2
$\begingroup$

I was wondering if any one has had success merging pre-generated reference index files for BWA, bowtie2 from multiple references?

For a concrete example I have hg19 and a bunch of viral and bacterial genomes that all have been index separately but I want to do a single alignment to all of them. Is there a way I could say combine hg19, virus_A, virus_D and bacteria_X into a single reference without having to regenerate the indexes? Since the problem could get combinatorial quickly and indexing anything with hg19 is time consuming I was wondering if there was a way around it.

$\endgroup$
10
  • $\begingroup$ this will likely depend on the tool no? i doubt there is a generic why to merge indexes? for BLAST, there is blastdb_aliastool $\endgroup$ Apr 4 '19 at 15:34
  • 1
    $\begingroup$ Yeah, I would split this into separate questions for each tool. I don't actually know how the blast index works, but I would be very surprised if the same approach will work for all three tools. It might be the same for bowtie and bwa, but blast will likely be different. $\endgroup$
    – terdon
    Apr 4 '19 at 16:09
  • $\begingroup$ Agreed I wasn't looking for a generic tools but the question does read that way. I was looking for a way for each of those tools. Thanks @Chris_Rands ill check out blastdb_aliastool $\endgroup$
    – Bioathlete
    Apr 4 '19 at 19:59
  • $\begingroup$ What would you be aligning? Short reads? Why would you want to align the same data against three so different genomes? Surely the time you would save by not re-indexing would be lost again while wasting time aligning to the wrong genome, right? $\endgroup$
    – terdon
    Apr 5 '19 at 8:34
  • $\begingroup$ @terdon I imagine they have metagenomic data or similar where the sample composition is not known a priori (reads could be 90% human, 8% E. coli etc.) $\endgroup$ Apr 5 '19 at 12:12

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.