Integration of expression QTLs with fine mapping via SuSiE

Area of Science:

• Genetics
• Statistical Genetics
• Bioinformatics

Background:

• Genome-wide association studies (GWASs) identify genetic variants linked to complex traits but struggle with pinpointing causal variants due to linkage disequilibrium (LD).
• Fine-mapping methods aim to assign causality probabilities to candidate variants by considering LD patterns.
• Integrating functional information, such as expression quantitative trait loci (eQTLs), can refine genetic fine-mapping.

Purpose of the Study:

• To introduce SuSiE2, a novel statistical framework that enhances genetic fine-mapping by incorporating eQTL information.
• To improve the accuracy and efficiency of identifying causal variants in GWASs.
• To leverage the Sum of Single-Effects (SuSiE) regression model for integrated eQTL and trait fine-mapping.

Main Methods:

• Developed SuSiE2, a method connecting two SuSiE models: one for eQTL analysis and another for trait fine-mapping.
• Computed posterior inclusion probabilities (PIPs) from an eQTL-based SuSiE model using gene expression levels.
• Utilized these PIPs as prior inclusion probabilities for risk variants in a trait-based SuSiE model.

Main Results:

• SuSiE2 demonstrated improved detection rates of causal single nucleotide polymorphisms (SNPs) compared to standard methods.
• The method reduced the average size of credible sets, enhancing precision in variant localization.
• Evaluations using simulations and real-world data (Alzheimer's disease, BMI) confirmed SuSiE2's superior performance in power, coverage, and precision.

Conclusions:

• SuSiE2 effectively integrates eQTL data into genetic fine-mapping, improving causal variant identification.
• The framework offers enhanced power and precision for fine-mapping complex traits, outperforming existing multi-trait methods.
• SuSiE2 provides a robust approach for dissecting the genetic architecture of complex diseases using GWAS and eQTL data.