Le Son Tran

CALFP-MHC improves peptide-MHC binding prediction by using chemically grounded fingerprints and contrastive learning, achieving high AUC and PPV across diverse alleles and class imbalance.

This study identifies ER O-glycosylation by GALNTs in synovial fibroblasts as a driver of cartilage degradation in arthritis, with Calnexin as a potential therapeutic target.

This paper uses supervised non-negative matrix factorization on cfDNA fragmentomic features to improve early cancer detection, achieving high AUC scores in both pan-cancer and HCC-specific classifications.

This paper develops a machine learning model that integrates TCRβ and pHLA binding data to improve neoantigen prioritization in colorectal cancer, showing better performance than existing tools.

The paper develops a model combining tumor-specific methylation atlas and cfDNA features to improve tissue-of-origin detection in low-depth cfDNA samples.