Predicting causal genes from psychiatric genome-wide association studies using high-level etiological knowledge

Area of Science:

• Genetics
• Psychiatric Research
• Bioinformatics

Background:

• Genome-wide association studies (GWAS) identify numerous genetic loci linked to psychiatric traits.
• Identifying the specific causal genes within these loci remains a significant challenge, impeding clinical applications.

Purpose of the Study:

• To introduce a novel method, Psychiatric Omnilocus Prioritization Score (PsyOPS), for prioritizing causal genes at psychiatric GWAS loci.
• To demonstrate PsyOPS's effectiveness and compare it with existing gene prioritization approaches.

Main Methods:

• PsyOPS utilizes three binary features reflecting key assumptions about psychiatric gene etiology: mutational constraint, brain-specific expression, and overlap with neurodevelopmental disease genes.
• Performance was evaluated against a feature-rich, domain-agnostic method.

Main Results:

• PsyOPS achieves state-of-the-art performance in causal gene prioritization for psychiatric GWAS loci, comparable to more complex methods.
• Genes prioritized by PsyOPS show stronger evidence of direct regulation by GWAS variants via DNA looping and quantitative trait locus (QTL) data.
• The method successfully prioritized genes like GABBR1 for schizophrenia, even when located far from the lead GWAS variant.

Conclusions:

• PsyOPS offers a highly effective and simplified approach to causal gene discovery in psychiatric genetics.
• Integrating high-level etiological knowledge significantly enhances the accuracy of gene prioritization.
• The prioritized genes serve as valuable candidates for experimental validation in psychiatric research.