Haoran Ma

This paper introduces gene-family encoding to improve domain-adapted single-cell language models, showing better performance in embedding-based and generative tasks on gastric-cancer data.

This study uses Perturb-seq to analyze pharmacogenomic interactions in gastric cancer, revealing gene-drug relationships and resistance mechanisms.

This study integrates spatial transcriptomics and single-cell data from gastric cancer samples to reveal tumor ecosystem dynamics, evolutionary trajectories, and immune microenvironment features.

This study identifies transcriptomic and immune features of gastric cancer peritoneal metastases, highlighting predictive markers and therapeutic targets in the tumor microenvironment.