Current challenges in GWAS integration and fine-mapping for variant interpretation

Area of Science:

• Genetics
• Genomics
• Bioinformatics

Background:

• Genome-wide association studies (GWAS) have identified numerous loci linked to complex traits and diseases over the past 20 years.
• However, the precise biological mechanisms and gene regulatory roles of these identified loci remain largely uncharacterized.
• This knowledge gap hinders the development of targeted therapeutics and accurate disease risk prediction.

Purpose of the Study:

• To address the challenges in prioritizing genetic variants from GWAS for functional follow-up experiments.
• To analyze current variant prioritization methods, including statistical fine-mapping and deep learning approaches.
• To propose a comprehensive strategy for resolving GWAS loci into high-confidence variants for mechanistic studies.

Main Methods:

• Analysis of limitations in data sharing, harmonization, and statistical/functional fine-mapping techniques.
• Evaluation of the utility of deep learning frameworks in conjunction with traditional statistical genetics.
• Development of a multi-modal approach integrating diverse data types for variant prioritization.

Main Results:

• Identified significant challenges in current GWAS variant prioritization, including data accessibility and methodological limitations.
• Highlighted the ambiguity in the added value of deep learning for variant effect prediction.
• Proposed a multi-modal strategy to enhance the accuracy and scalability of variant prioritization.

Conclusions:

• A multi-modal approach is crucial for effectively prioritizing GWAS-identified variants for functional exploration.
• Improved data sharing, harmonization of summary statistics, and in-sample linkage disequilibrium (LD) data are essential.
• Realizing the full potential of GWAS requires robust and scalable methods for causal variant identification.