genotype_for_sample()

[pnrobinson]

I am not sure why I am seeing this method for Variant, that seems like the wrong place to retrieve the genotype for a sample specified in the argument?

[ielis]

It indeed looks a bit weird. The exact details of Variant API still need to be worked out.

We need to figure out how to manage 2 states: individual variant and cohort variant.

The individual variant represents data available right after loading phenopacket. We have the VariantCoordinates and genotype. The genotype is usually non-homozygous reference.

The cohort variant is an aggregation of the data of all cohort members. It includes the VariantCoordinates, functional annotation(s), and mapping from sample_id -> genotype. In contrast with individual variant, here the genotype can be ./. or 0/0.

I think we can model the Variant in a similar fashion to VCF format, where we have 3 levels of information for each variant:

• variant coordinates (CHROM, POS, ID, REF, ALT)
• variant-related data such as functional annotations (INFO)
• genotype-related data - mapping from sample_id to genotype

All this said, we may need to discuss this further before starting the actual work.

[pnrobinson]

@lnrekerle is this still active or can the issue be closed?

[lnrekerle]

I think this can be closed since this method no longer exists.