Quantitative omnigenic model discovers interpretable genome-wide associations.

Area of Science:

• Genetics and Genomics
• Systems Biology
• Statistical Modeling

Background:

• Genome-wide association studies (GWAS) identify numerous loci for complex traits, each explaining small variance fractions.
• The omnigenic model proposes distant loci contribute via regulatory network propagation.
• Existing models struggle to integrate network topology with genomic data for complex trait analysis.

Purpose of the Study:

• To formalize the omnigenic model into a statistical framework: the quantitative omnigenic model (QOM).
• To improve gene expression prediction and understand trait heritability through regulatory networks.
• To assess the role of transcriptional and non-transcriptional effects in complex traits.

Main Methods:

• Developed the quantitative omnigenic model (QOM) integrating regulatory network topology with genomic data.
• Applied QOM to gene expression traits in yeast.
• Estimated cis- and trans- heritable variance, and effect propagation orders.

Main Results:

• QOM achieved comparable gene expression prediction to GWAS with significantly fewer parameters.
• QOM identified candidate causal chains of effect propagation through regulatory networks.
• QOM accurately accounted for gene expression covariance structure and differentiated transcriptional from non-transcriptional effects.

Conclusions:

• QOM provides a powerful statistical framework for analyzing complex traits by incorporating regulatory network information.
• The model enhances prediction accuracy and biological interpretability of genetic effects.
• QOM is relevant for systems biology, network quality assessment, and understanding diverse effect propagation.