fullsniplings: Full-sibling inference from biallelic markers (like SNPs)

In eriqande/fullsniplings: Full sibling reconstruction from SNP data

Description

fulsniplings is a package in development. It implements some new statistical perspectives and optimized computational tools to do MCMC over the posterior distribution of pedigrees given genetic data.

Details

Geared toward datasets with many individuals and about 100 SNPs which are treated as if they are unlinked. (While that is clearly false, it might be reasonable). The name derives from the fact that the idea for the approach taken here came to me when I was working on sibling inference issues. It became clear that being able to compute the posterior predictive distribution of genotypes conditional on observed sibling groups would be a powerful way forward in the sibling inference problem. I will eventually generalize that to more general pedigree inference, but for now just want to implement the easy full-sibling case as a prototypical exercise.

It really is just in development at the moment

eriqande/fullsniplings documentation built on May 16, 2019, 8:45 a.m.