GenePy - a score for estimating gene pathogenicity in individuals using next-generation sequencing data.

Area of Science:

• Genomics
• Bioinformatics
• Computational Biology

Background:

• Next-generation sequencing (NGS) has advanced rare disease diagnosis but is limited in understanding common disease aetiology.
• Binary genetic models fail to capture the cumulative effect of multiple variants in complex diseases.
• A novel gene-level scoring system, GenePy, is needed to interpret complex genetic contributions to disease.

Purpose of the Study:

• To introduce GenePy, a novel gene-level scoring system for per-individual analysis of NGS data.
• To transform NGS data interpretation from variant-level to gene-level for complex phenotypes.
• To facilitate integration with machine learning and network approaches for disease investigation.

Main Methods:

• Developed GenePy, a scoring system integrating population allele frequency, individual zygosity, and deleteriousness metrics.
• Calculated GenePy scores for variants and aggregated them into a single gene score per individual.
• Applied GenePy to whole-exome sequencing data from 508 individuals, generating ~14,000 gene scores across 16 deleteriousness metrics.

Main Results:

• GenePy scores were generated for ~14,000 genes across 508 individuals, corrected for gene length.
• Individuals with multiple rare, deleterious mutations can accumulate very high GenePy scores.
• GenePy significantly outperformed a standard association tool in discriminating genes linked to three complex diseases (p=1.37×10⁻⁴ vs p=0.003).

Conclusions:

• GenePy provides intuitive per-gene, per-individual scores for assessing genetic variation.
• GenePy surpasses current best-practice tools for integrating common and rare genetic variations.
• GenePy scores are suitable for integration with transcriptomic and proteomic data for comprehensive analysis.