Lurong Pan

AINN-Express predicts VHH antibody expression using only sequence data, achieving high ROC-AUC scores in both known and new programs without requiring structural data or task-specific training.

The paper shows that frozen embeddings from a protein foundation model outperform a task-specific model in ranking antibody-antigen pairs by binding affinity.

This study uses AI to design mutation-resistant antibodies effective against multiple SARS-CoV-2 strains, validated experimentally against Omicron variants.

This paper develops an in-silico drug repurposing pipeline for epilepsy using deep learning and structure-based methods, identifying potential drugs with high binding affinity for epilepsy-related targets.