When calling population CNVs, I would use different tools (e.g., Lumpy. Delly, and Manta) to call SVs for each sample, and thus I will get three VCF files for each sample. Then, for each sample and each VCF, I will filter for the high quality variants. After, high quality variants from all above three algorithms were aggregated for each sample and now I will get one VCF file for each sample. Last, I will merge all the individual VCF files into one VCF file using SURVIVOR merge. Through above procedures, I think the breakpoints of variants will become coarse. Does this affect the genotype results (genotype the merge sites in all samples using svtyper) when I do not have a precise breakpoint?
Dear @brentp
When calling population CNVs, I would use different tools (e.g., Lumpy. Delly, and Manta) to call SVs for each sample, and thus I will get three VCF files for each sample. Then, for each sample and each VCF, I will filter for the high quality variants. After, high quality variants from all above three algorithms were aggregated for each sample and now I will get one VCF file for each sample. Last, I will merge all the individual VCF files into one VCF file using
SURVIVOR merge. Through above procedures, I think the breakpoints of variants will become coarse. Does this affect the genotype results (genotype the merge sites in all samples using svtyper) when I do not have a precise breakpoint?Sincerely, Zhuqing