Manuel Corpas

Benchmarking large language models for ACMG/AMP variant interpretation and variant calling

This paper benchmarks large language models for variant interpretation and calling by evaluating safety, validity, and reproducibility rather than accuracy alone.

Trustworthy agentic genomics through versioned skill libraries

This paper shows that trustworthy agentic genome interpretation requires executing validated skills as code rather than relying on models to reason or retrieve information, ensuring deterministic and auditable results.

Genomebook: Mendelian inheritance as a structured parameterisation layer for LLM agent populations

This paper introduces Genomebook, a system that encodes genetic inheritance in LLM agents, showing trait changes over generations under selective pressures.

The Peruvian Genome Project: expanding the global pool of genome diversity from South America

The Peruvian Genome Project collects and analyzes genomic data from native and mestizo populations in Peru to address underrepresentation in global genome databases and uncover genetic adaptations to local environments.

Low-coverage whole genome sequencing for a highly selective cohort of severe COVID-19 patients

This study uses low-coverage whole genome sequencing and imputation to analyze genetic variants in severe COVID-19 patients, showing high imputation accuracy and insights into clinical phenotypes.

Implementation of Individualised Polygenic Risk Score Analysis: A Test Case of a Family of Four

The paper calculates polygenic risk scores for a family using whole genome sequencing and 1000 Genomes Project data to demonstrate individual-level PRS implementation.

PGP-UK: a research and citizen science hybrid project in support of personalized medicine

PGP-UK explores citizen science in genomics by recruiting participants, sharing data, and developing tools to advance personalized medicine.

A Collection Of 2,280 Public Domain (CC0) Curated Genotypes

The paper curates a public domain genome dataset of 2,280 individuals from 23andMe under CC0 license for research use.

10 Simple Rules for Sharing Human Genomic Data

The paper outlines 10 rules for sharing human genomic data to maximize research value while protecting participant privacy.