A Bayesian fine-mapping model using a continuous global-local shrinkage prior with applications in prostate cancer...

Area of Science:

• Genetics
• Bioinformatics
• Computational Biology

Background:

• Fine-mapping aims to pinpoint genetic variants linked to complex traits and diseases.
• Current methods using discrete priors and fixed maximum variant counts can be suboptimal.

Purpose of the Study:

• Introduce h2-D2, a Bayesian fine-mapping method employing a continuous global-local shrinkage prior.
• Develop a method for defining credible sets of causal variants within continuous prior frameworks.

Main Methods:

• Utilized a novel continuous global-local shrinkage prior for Bayesian fine-mapping.
• Developed a new approach for constructing credible sets of causal variants.
• Performed simulation studies to compare h2-D2 with existing methods (SuSiE, FINEMAP).

Main Results:

• h2-D2 demonstrated superior performance in identifying causal variants and estimating their effect sizes compared to state-of-the-art methods.
• Application to prostate cancer data identified novel causal variants.
• Inferred 369 target genes and significantly over-represented pathways, offering insights into prostate cancer mechanisms.

Conclusions:

• h2-D2 offers an improved approach to genetic fine-mapping for complex traits and diseases.
• The method provides valuable insights into the genetic architecture of prostate cancer, identifying potential therapeutic targets.