Jie Tang

InstructPLM aligns a protein language model with structure instructions to generate diverse proteins, improving performance over existing models in sequence recovery and activity.

A 100-billion-parameter protein language model trained with masked and autoregressive objectives together, topping prior results on 15 of 18 benchmarks and powering a fast structure predictor.

This paper introduces ESMPair, a method using protein language models to improve interolog identification, leading to better protein complex structure prediction than AlphaFold-Multimer.