Identifying individuals with rare disease variants by inferring shared ancestral haplotypes from SNP array data

Abstract

We describe FoundHaplo, a novel identity-by-descent algorithm designed to identify individuals with known, untyped, disease-causing variants using only SNP array data. FoundHaplo leverages knowledge of shared disease haplotypes for inherited disease-causing variants to identify individuals who share the disease haplotype and are, therefore, likely to carry the rare (MAF<0.01) variant. We performed a simulation study to evaluate the performance of FoundHaplo across 33 known disease-harbouring loci. We demonstrated the ability of FoundHaplo to infer the presence of two rare (MAF<0.01) pathogenic variants,SCN1Bc.363C>G (p.Cys121Trp) andWWOXc.49G>A (p.E17K), which can cause mild dominant and severe recessive epilepsy respectively, in two large cohorts including 1,573 individuals with epilepsy from the Epi25 cohort and 468,481 individuals from the UK Biobank. We demonstrate that FoundHaplo performs substantially better at inferring the presence of these variants than existing genome-wide imputation approaches. FoundHaplo is a valuable, low-cost screening tool that can be applied to search SNP genotyping array data for disease-causing variants with known founder effects based on shared disease haplotypes. FoundHaplo is available athttps://github.com/bahlolab/FoundHaplo.