FID Definition

[Kristen James]

I am performing QC on a cohort of ~300 individuals who were genotyped using the Illumina Global Screening array. Using the identical by descent --genome command I identified a handful of first degree relatives (pi-hat ~= 0.5) and now want to incorporate this information in the family file. My question is, should unrelated parents have the same family ID (FID)?

For example, MomA and DadA have ChildA and ChildB. Should all individuals in this cluster have the same FID even though MomA and DadA are genetically unrelated? 

A broader question, how is the FID versus the maternal or paternal ID used in downstream analyses?

Thank You,

Kristen

[Christopher Chang]

For most plink functions, it doesn't matter what ID scheme you use; just pick a scheme that helps you keep everything organized.  (The main exception is if you're planning to perform a family-based analysis.)