networkGWAS: a network-based approach to discover genetic associations

Area of Science:

• Genetics
• Bioinformatics
• Computational Biology

Background:

• Genome-wide association studies (GWAS) identify genetic variants linked to complex traits, but often explain limited phenotypic variation.
• Existing network-based GWAS methods face challenges with vast search spaces and multiple testing, leading to missed associations or false positives.

Purpose of the Study:

• To develop a computationally efficient and statistically sound method for network-based genome-wide association studies.
• To integrate biological network information with gene-based GWAS to improve the detection of trait-associated genes.

Main Methods:

• Proposed networkGWAS, a novel approach utilizing mixed models and neighborhood aggregation for network-based GWAS.
• Implemented population structure correction and employed circular and degree-preserving network permutations for well-calibrated P-values.

Main Results:

• networkGWAS successfully identified known associations in synthetic phenotypes.
• The method detected known and novel genes associated with phenotypes in Saccharomyces cerevisiae and Homo sapiens.
• Demonstrated the ability to systematically combine gene-based GWAS with biological network data.

Conclusions:

• networkGWAS offers a statistically robust and computationally efficient solution for network-based GWAS.
• The approach enhances the discovery of genetic associations by integrating biological network information.
• Enables more comprehensive analysis of complex traits by combining GWAS with network biology.