Chengzhong Ye

This paper introduces GPN-Star, a phylogeny-informed genomic language model that improves variant effect prediction across species and evolutionary timescales.

GPN-MSA is a DNA language model that predicts variant effects across the human genome using whole-genome alignments, achieving strong performance on deleteriousness prediction for coding and non-coding variants.

The paper introduces epiTree, a pipeline using tree-based models to detect higher-order epistatic interactions in genomic data, validated on UK Biobank data for red hair and MS prediction.

This paper introduces a cross-protein transfer learning framework that improves disease variant prediction using deep mutational scanning data from few proteins, achieving better specificity than existing models.

The paper introduces a deep learning framework to impute surface protein abundances from scRNA-seq data using transfer learning.

This paper improves scRNA-seq data quality using transfer learning by combining a deep autoencoder with a Bayesian model to extract gene-gene relationships across diverse datasets.

DECENT improves differential expression analysis in scRNA-seq by modeling dropout, outperforming existing methods on simulated and real data without requiring spike-ins.