problem with file .fam

[rania Abdelhedi]

Hi everyone,
I am working on a whole-genome sequencing (WGS) project and I am currently converting my VCF files to PLINK binary format (BED/BIM/FAM).
I’ve noticed that when I generate the .fam file, the sex, phenotype, and pedigree columns are filled with default values (0 for IDs/sex, -9 for phenotypes). I understand that PLINK generally expects these as external inputs, but I wanted to confirm: does PLINK have any built-in functionality to automatically infer or populate these fields from WGS data, or is it strictly required to manually provide this metadata via flags like --update-sex or --update-pheno?
Are there any standard pipelines or best practices for handling metadata integration in large-scale WGS studies?

[Christopher Chang]

It's usually fine to leave pedigree and phenotype unset in .fam files; phenotype values are often stored in a separate file, loaded with --pheno when needed.

However, sex should be populated, especially if you have any chrX/chrY/chrM data.  It is best to provide this information if you have it, but PLINK has an --impute-sex command you can use in a pinch (note that PLINK 2.0 --impute-sex has better guardrails than PLINK 1.9's implementation).