Inferring compound heterozygosity from large-scale exome sequencing data.

Area of Science:

• Genetics
• Genomic Medicine
• Computational Biology

Background:

• Recessive diseases require two damaging genetic variants, one on each chromosome copy (in trans), for manifestation.
• Distinguishing between cis and trans variants is essential for accurate genetic diagnosis but challenging clinically.
• Current methods for variant phasing are often limited in routine clinical application.

Purpose of the Study:

• To develop and validate a computational strategy for inferring the phase of rare variant pairs within genes.
• To leverage population-scale exome data for accurate variant phasing.
• To provide a resource to aid in the interpretation of compound heterozygous variants for recessive disease diagnosis.

Main Methods:

• Inferred phase for rare variant pairs using genotypes from the Genome Aggregation Database (gnomAD v2, n=125,748 exomes).
• Leveraged large-scale exome sequencing data to estimate the phase of variants.
• Validated the phasing accuracy using trio data and patient cohorts with Mendelian conditions.

Main Results:

• The developed strategy accurately estimates variant phase with 96% accuracy.
• Phasing accuracy was confirmed in both family-based (trio) data and patient cohorts.
• A public resource of phasing estimates and counts of rare variants in trans per gene was generated.

Conclusions:

• The novel phasing strategy significantly improves the ability to interpret compound heterozygous variants.
• This approach enhances the diagnostic yield for rare genetic disorders.
• The public resource facilitates the interpretation of complex genetic findings in recessive disease.