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

Area of Science:

• Genetics
• Bioinformatics
• Computational Biology

Background:

• Identifying rare disease-causing variants is crucial for genetic research and diagnostics.
• Existing methods like genome-wide imputation have limitations in detecting very rare variants.

Purpose of the Study:

• To introduce FoundHaplo, a novel algorithm for screening untyped, rare disease-causing variants using SNP array data.
• To evaluate FoundHaplo's performance in detecting specific pathogenic variants in human cohorts.

Main Methods:

• Developed FoundHaplo, an identity-by-descent algorithm leveraging shared disease haplotypes.
• Applied FoundHaplo to simulate data and real-world cohorts (Epi25, UK Biobank) to detect rare variants (MAF ≤ 0.01%).
• Compared FoundHaplo's sensitivity against genome-wide imputation.

Main Results:

• FoundHaplo successfully inferred the presence of rare pathogenic variants (SCN1B and WWOX) in human cohorts.
• The algorithm demonstrated substantially higher sensitivity compared to existing genome-wide imputation methods.
• Simulation studies confirmed FoundHaplo's performance across 33 disease-harboring loci.

Conclusions:

• FoundHaplo is an effective screening tool for identifying rare variants with founder effects using SNP genotyping data.
• The algorithm has broad applicability in human disease genetics, nonhuman applications, and quantitative trait analysis.
• FoundHaplo offers a valuable approach for discovering genetic variants underlying various traits.