Shane O’Connell

DNT: Diploid Genomic Foundation Model

The paper introduces a diploid genomic foundation model to represent diploid genotypes in genomic language models, improving variant discrimination in compound-heterozygous cases.

GFMBench-API: A Standardized Interface for Benchmarking Genomic Foundation Models

Wrap a genomic foundation model's inference methods once to score it on supervised and zero-shot GUE, VariantBenchmarks, BEND, TraitGym and ClinVar tasks; code is promised upon acceptance.

Suicidality phenotypes reflect both shared and distinct genetic factors

This study identifies shared and distinct genetic factors underlying suicidality phenotypes using GWAS and structural equation modelling.

Deriving Mendelian Randomization-based Causal Networks of Brain Imaging Phenotypes and Bipolar Disorder

This study uses Mendelian Randomization to identify causal relationships between neuroimaging phenotypes and bipolar disorder, finding that white matter tract variations have larger causal effects on BD.

Convex approaches to isolate the shared and distinct genetic structures of subphenotypes in heterogeneous complex traits

The paper introduces Clorinn, a convex optimization method to decompose genetic variant associations into shared and distinct components across complex traits, showing improved performance in simulations and application to GWAS data.