VCF file query

8 views

Skip to first unread message

huma rafique

unread,

Aug 1, 2026, 10:54:57 PMAug 1

to fastsimcoal2

I am planning to compare reciprocal gene-flow models among different goat populations using fastsimcoal2.

Unfortunately, the only combined VCF available has been filtered at MAF > 0.005, phased with Beagle, and contains imputed genotypes/markers (IMP, DR2, and DS fields). The original unfiltered, non-imputed VCF or raw gVCFs are no longer available.

Would it be methodologically acceptable to construct a folded multidimensional SFS from this VCF and compare alternative gene-flow models? Could removing IMP-flagged sites and retaining only high-DR2 variants make the analysis reliable, or would the prior MAF filtering and imputation make the demographic inference too biased?

Any recommendations for handling this situation would be greatly appreciated.

Best regards,
Huma Ghazal

Reply all

Reply to author

Forward

0 new messages