Zichen Zhang

Biological rationales from language models enable leakage-resistant forecasts of target-indication success

PRIORITI uses language models to predict target-indication success in drug discovery, achieving high accuracy and robustness across different benchmarks.

TigerAI: An AI-powered genetic evidence platform to support clinical development

Domain-grounded prompting lets GPT-5 produce genetic evidence for 13,022 drug target-indication pairs that predicts clinical success as well as expert curation, with wider coverage.

An atlas of genetic effects on the monocyte methylome across European and African populations

The paper constructs a multi-ancestry methylome atlas of purified monocytes, identifies population-specific methylation quantitative trait loci, and performs methylome-wide association studies for 41 traits, revealing novel associations for type 2 diabetes.

Large-scale integration of omics and electronic health records to identify potential risk protein biomarkers and therapeutic drugs for cancer prevention

This study integrates large-scale omics and electronic health records to identify risk proteins and drugs for cancer prevention, finding 101 novel proteins and three drugs associated with reduced colorectal cancer risk.

SUMMIT-FA: A new resource for improved transcriptome imputation using functional annotations

SUMMIT-FA improves gene expression prediction accuracy using functional annotations and eQTL data, enhancing gene-trait association detection in GWAS.

Large-scale imputation models for multi-ancestry proteome-wide association analysis

The paper introduces BLISS, a framework for training proteome-wide association models using pQTL summary statistics, enabling multi-ancestry analysis of protein-phenotype associations.

MIMOSA: A resource consisting of improved methylome imputation models increases power to identify DNA methylation-phenotype associations

MIMOSA improves DNA methylation prediction and MWAS power by integrating summary-level mQTL data, identifying more CpG-phenotype associations in whole blood.

SUMMIT: An integrative approach for better transcriptomic data imputation improves causal gene identification

SUMMIT improves transcriptomic data imputation and TWAS power by using large eQTL summary datasets, identifying 11 causal genes for COVID-19 severity.