The flashfm approach for fine-mapping multiple quantitative traits.

Area of Science:

• Genetics
• Statistical genetics
• Bioinformatics

Background:

• Fine-mapping genetic association signals is crucial for identifying causal variants.
• Leveraging information across multiple quantitative traits can enhance fine-mapping accuracy and resolution compared to single-trait approaches.

Purpose of the Study:

• To introduce flashfm (flexible and shared information fine-mapping), a novel Bayesian method for joint fine-mapping of multiple quantitative traits.
• To evaluate flashfm's performance in simulations and a real-world cardiometabolic trait dataset.

Main Methods:

• flashfm utilizes summary statistics within a Bayesian framework to jointly fine-map signals for multiple traits.
• The method incorporates prior model probabilities favoring shared causal variants and accommodates missing trait data and related individuals.

Main Results:

• Simulations show flashfm performs comparably to single-trait fine-mapping when traits lack shared causal variants.
• When traits share causal variants, flashfm reduces the number of potential causal variants by up to 30% compared to single-trait methods.
• Application to 33 cardiometabolic traits in a Ugandan cohort yielded a 20% reduction in potential causal variants.

Conclusions:

• flashfm is a computationally efficient tool for joint fine-mapping of multiple traits using summary statistics.
• The method effectively leverages shared genetic architecture across traits to improve fine-mapping resolution.
• flashfm can be readily applied to large-scale genetic studies with up to six traits.