Dissecting the genetic overlap between three complex phenotypes with trivariate MiXeR

Area of Science:

• Genetics and Bioinformatics
• Computational Biology
• Human Complex Traits

Background:

• Comorbidities represent a growing global health concern, with increasing evidence pointing to shared genetic factors underlying complex traits and disorders.
• Existing methods, like the bivariate causal mixture model (MiXeR), can quantify genetic overlap between two phenotypes but are insufficient for disentangling the complex genetic architecture of multimorbidity involving three or more conditions.

Approach:

• Development and validation of the trivariate MiXeR, a novel computational tool designed to quantify the polygenic overlap among three distinct phenotypes.
• The trivariate MiXeR utilizes summary statistics from genome-wide association studies (GWAS) to disentangle complex patterns of genetic overlap.
• The tool's performance was assessed through simulations and applied to real GWAS data to estimate genetic overlap proportions between multiple traits.

Key Points:

• The trivariate MiXeR reliably reconstructs diverse patterns of genetic overlap in simulation studies.
• Application to real GWAS data reveals complex genetic overlap patterns between human traits and diseases that are undetectable with bivariate analyses.
• The tool provides a method to estimate the proportions of genetic overlap among three phenotypes, offering deeper insights into their shared etiology.

Conclusions:

• The trivariate MiXeR significantly enhances the ability to characterize multimorbidity by quantifying genetic overlap between three phenotypes.
• This advancement contributes to a more comprehensive understanding of the etiology of complex phenotypes and their interrelationships.
• The findings may aid in dissecting comorbidity patterns and elucidating their underlying biological mechanisms.