windowsize option not working

[Tgrandis]

I tried to extract all reads aligning to a specific region from a bam file. I tried to set the window size larger than the default = 5, but no matter what value I entered larger or smaller, it returned to default.

[tobiasrausch]

Thanks for the bug report. That's fixed in the github head now. The window length specifies the region around -p that needs to be spanned by the input reads. You can also trim the reads to that window length, i.e.:

alfred consensus -w 20 -r -p chr4:500500 input.bam

[Tgrandis]

Hello Tobias

Thanks for the very quick response. The -w option in consensus is now functional... but it functions in a different way from what I had expected. I had hoped that a larger windowsize would widen the region from which reads would be extracted from the bam alignment. It turns out though that it now extracts only those reads that completely overlap the chosen window, rather than overlap anywhere within the window.

Is there any other way to extract reads from a larger region in the bam file and display them? I can view regions in IGV, but I cannot manipulate them in IGV. I would like to be able to eliminate erroneously aligned reads, sort the reads by genotype at SNPs and ultimately also connect reads with their paired-end reads in windows of 500 to maybe 1000 bp long. All I need is the fasta sequences that map onto the selected region of the bam file very similar to what the consensus command does in alfred. An iterative consensus command would actually do the job, stepwise extracting all the reads mapping to position x, x+5, x+10, .... x+n , eliminating all duplicates and aligning them back with the proper position.

Best Regards,

Hugo

---

Hugo A. Volkaert Center for Agricultural Biotechnology Kasetsart University Kamphaengsaen Campus Kamphaengsaen, Nakhonpathom Thailand 73140 phone: +66 34353217 fax: +66 34353222

-------- Original Message --------

    SUBJECT:
    Re: [tobiasrausch/alfred] windowsize option not working (Issue #93)

    DATE:
    2022-10-27 01:31

    FROM:
    Tobias Rausch ***@***.***>

    TO:
    tobiasrausch/alfred ***@***.***>

Thanks for the bug report. That's fixed in the github head now. The window length specifies the region around -p that needs to be spanned by the input reads. You can also trim the reads to that window length, i.e.:

alfred consensus -w 20 -r -p chr4:500500 input.bam

-- Reply to this email directly, view it on GitHub [1], or unsubscribe [2]. You are receiving this because you authored the thread.Message ID: @.***>

Links:

[1] https://github.com/tobiasrausch/alfred/issues/93#issuecomment-1292439000 [2] https://github.com/notifications/unsubscribe-auth/ADIOC7NKIVQS3LZPL6ZZ3RDWFF2HPANCNFSM6AAAAAARO7ZTUM