Andrew B. Conley

This paper trains a deep learning model on multiomics data to predict metabolic dynamics in synthetic biology, achieving high accuracy in predicting product and metabolite levels.

Rye infers genetic ancestry at biobank scale with high accuracy and 50x faster runtime than ADMIXTURE on UK Biobank data.

The paper assesses the utility of genetic ancestry for pharmacogenomic risk stratification compared to self-identified race/ethnicity in a U.S. cohort.

The paper identifies admixture-enabled selection in Latin American populations, finding evidence of African ancestry enrichment at immune-related loci and polygenic selection for traits related to immunity and inflammation.

The paper analyzes Native American ancestry in three US populations, revealing distinct patterns linked to historical migration and settlement.

The paper identifies phenotypic cues and genetic architecture underlying ancestry-based assortative mating in admixed Latin American populations using genome-wide haplotype analysis.