Paul J. Newcombe

This paper introduces mJAM, a multi-population fine-mapping method that improves credible set construction by incorporating population-specific LD structures and demonstrates its effectiveness in real GWAS data.

Flashfm improves fine-mapping of multiple quantitative traits by sharing information between traits, reducing potential causal variants by 20% in a Ugandan cohort.

This paper introduces a Bayesian variable selection algorithm for Mendelian randomization that identifies genetic variants with minimal pleiotropy, showing significant risk-increasing effects of blood pressure on coronary heart disease.

This paper introduces hJAM, a hierarchical model that improves causal inference in Mendelian randomization and transcriptome-wide association studies by incorporating prior information and accounting for SNP correlations.

The paper introduces a Bayesian method to fine map direct chromatin contacts in capture Hi-C data by identifying a subset of true contacts from apparent long runs of co-contacting fragments.