Simplifying causal gene identification in GWAS loci

Area of Science:

• Genetics
• Bioinformatics
• Computational Biology

Background:

• Genome-wide association studies (GWAS) identify genetic variants linked to diseases, but pinpointing causal genes within these loci is difficult.
• Current gene prioritization tools often rely on complex models and may contain biases, necessitating simpler, more robust approaches.

Purpose of the Study:

• To develop a data-driven approach for prioritizing causal genes in GWAS loci.
• To create an accurate, interpretable, and computationally efficient gene prioritization tool.

Main Methods:

• Constructed a truth set of causal genes from 200 GWAS loci using a data-driven approach.
• Developed CALDERA, a gene prioritization tool employing a simple logistic regression model with four input features.
• Benchmarked CALDERA against existing methods (FLAMES, L2G, cS2G) on independent datasets.

Main Results:

• A simple logistic regression model performed comparably to a complex XGBoost model for gene prioritization.
• CALDERA achieved state-of-the-art performance, outperforming other methods in independent benchmarking.
• CALDERA successfully predicted 11,956 putative causal genes for 93 UK Biobank traits, potentially resolving up to 52% of loci.

Conclusions:

• CALDERA offers a powerful and efficient solution for prioritizing potentially causal genes in GWAS loci.
• The tool minimizes data processing requirements and provides well-calibrated, interpretable causal gene probabilities.
• CALDERA has the potential to significantly advance the interpretation of GWAS findings.