search

chhylp123 / hifiasm

Hifiasm: a haplotype-resolved assembler for accurate Hifi reads
MIT License
542 stars 87 forks source link

parental data recommendation #491

Open LHG-GG opened 1 year ago

LHG-GG commented 1 year ago

Hi @chhylp123,

I have both short reads (Illumina) and long reads (Pacbio HiFi) for parents. Which data do you recommend to use for hifiasm-trio mode? or merge short reads and long reads?

For yak count, whether I have multiple paired short read files (multiple R1 R2) or HiFi data (hifi1.fq, hifi2.fq, etc), I should use (https://github.com/chhylp123/hifiasm/issues/262) yak count -k31 -b37 -o mat.yak <(cat *.fq) <(cat *.fq)

Thank you.

chhylp123 commented 1 year ago

Either illumina or HiFi should be fine. Hifiasm should work with both illumina and HiFi as well. As for yak index, yes.

LHG-GG commented 1 year ago

Thank you @chhylp123. From your experience any suggestion on which is best suited (hifi or illumina) get assembly with less switch errors/hamming error with increased phased blocks or hifiasm does not benefit from parental read lenght?

chhylp123 commented 1 year ago

I feel like phasing with hifi or illumina should only have minor difference. But you can quickly have a try with the bin files of hifiasm using different types of parental data.

DustinSokolowski commented 3 months ago

Thank you for the helpful discussion. Would you expect that herro-corrrected ONT R.10 (called with Dorado V5) parental data would also correctly trio bin? My impression is that these data can be treated like hifi data but I would appreciate your opinion as well.

Thanks! Dustin