DecodeGenetics / graphtyper

Population-scale genotyping using pangenome graphs
http://dx.doi.org/10.1038/ng.3964
MIT License
167 stars 20 forks source link

All BNDs have no FORMAT:FT #127

Open tgong1 opened 1 year ago

tgong1 commented 1 year ago

Hi,

There are no FT tags in FORMAT fields for ALL BNDs. Do you have any idea why this happen? Any idea or suggestion is appreciated.

Thank you, Tingting

tgong1 commented 1 year ago

Hi, I also found all BNDs have only AGGREGATED models. Is this what expected to see? I have used Manta for calling and svimmer for SV merging.

Thank you, Tingting

jingydz commented 11 months ago

Hi, do you have solved this problem? I meet the same problem. normal SV type, such as DEL, INS, DUP, all have FT field but BND no FT field, before vcffillter -g FT=PASS image after vcffilter -g FT=PASS image

all sample genotype of BND changed from 0/0 or 0/1 into ".",that 's true?

tgong1 commented 11 months ago

Hi,

No, I haven't solve the problem of BND. So I have given up the method of filtering based on FT=PASS. I'm now using the PASS_ratio for filtering.

Best regards, Tingting

jingydz commented 11 months ago

BND:-f "( SVTYPE = BND & QD > 20 & ( ABHet > 0.30 | ABHet < 0 ) & ( AC / NUM_MERGED_SVS ) < 10 & PASS_AC > 0 & PASS_ratio > 0.1 )"

-f, --info-filter specifies a filter to apply to the info fields of records, removes alleles which do not pass the filter

-g, --genotype-filter specifies a filter to apply to the genotype fields of records

If you only use the PASS_ratio, then it only filter the BND records but not sample genotype? can it filter unpassed genotype?

tgong1 commented 11 months ago

@jingydz Yes, the filtering I'm using now is for SV sites only. While I have also tried on filtering BND based on GQ>40 (you have to set a threshold here), and other types based on FT=PASS. Hope that helps.

jingydz commented 11 months ago

Thank you again. Finally I used "( GQ > 20 ) & ( AD > 10 )" for BND and "FT=PASS" for other types.