We describe FoundHaplo, an identity-by-descent algorithm that can be used to screen untyped disease-causing variants using single nucleotide polymorphism (SNP) array data. FoundHaplo leverages knowledge of shared disease haplotypes for inherited variants to identify those who share the disease haplotype and are, therefore, likely to carry the rare [minor allele frequency (MAF) ≤ 0.01%] variant. We performed a simulation study to evaluate the performance of FoundHaplo across 33 disease-harbouring loci. FoundHaplo was used to infer the presence of two rare (MAF ≤ 0.01%) pathogenic variants, SCN1B c.363C>G (p.Cys121Trp) and WWOX c.49G>A (p.E17K), which can cause mild dominant and severe recessive epilepsy, respectively, in the Epi25 cohort and the UK Biobank. FoundHaplo demonstrated substantially better sensitivity at inferring the presence of these rare variants than existing genome-wide imputation. FoundHaplo is a valuable screening tool for searching disease-causing variants with known founder effects using only SNP genotyping data. It is also applicable to nonhuman applications and nondisease-causing traits, including rare-variant drivers of quantitative traits. The FoundHaplo algorithm is available at https://github.com/bahlolab/FoundHaplo (DOI:10.5281/zenodo.8058286)....

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

10.1093/nargab/lqaf033