search

twolinin / longphase

GNU General Public License v3.0
98 stars 6 forks source link

SNV with genotype 0/0 phased to 0|1 #66

Closed tgong1 closed 1 month ago

tgong1 commented 3 months ago

Hi,

Thank you for developing the useful tool! I have multiple SNV haplotype outputs as 0|1 or 1|0, while their input genotypes are 0/0. I'm wondering is this expected to see by longphase?

Thank you, Tingting

twolinin commented 3 months ago

Hi,

Thank you for reporting the issue. As a general rule, we currently do not phase variants with a genotype of 0/0. Could you please provide the command you used, the version of LongPhase, or the files you were working with to help us clarify this issue?

Thanks

tgong1 commented 3 months ago

The longphase version I used is 1.5.1. Interestingly, I found the genotype of 0/0 was not phased when phasing SNVs only, while phased when co-phasing SNV and SVs. I randomly checked three phased 0/0 SNVs and found there are SVs in the exactly same position with genotype as 0/1. I feel the SNV outputs may somehow related to SV inputs? Hope it can provide some help for your investigation. Again, Thank you for your time and help!

The commands used as following: longphase_linux-x64 phase \ -s ${SNV} \ -b $bam \ -r $REF \ -t 4 \ -o $OUT \ --ont \ --indels

longphase_linux-x64 phase \ -s ${SNV} \ --sv-file ${SV} \ -b $bam \ -r $REF \ -t 8 \ -o $OUT \ --ont \ --indels

ythuang0522 commented 3 months ago

Hi @tgong1 Yes. That was a previous issue when different types of variants were overlapped (not only SVs, modifications can also overlap), which has been resolved in v1.7. We also encourage you to upgrade to the latest version for higher phasing accuracy.

ythuang0522 commented 3 months ago

Sorry. We just found the code was not merged into v1.7. Will release a hotfix version within a week or so.

ythuang0522 commented 2 months ago

Hi @tgong1 We have released v1.7.1 which should have resolved this issue. Please try again.