Guolin Ke

UFConf improves protein conformation sampling using a diffusion model based on amino acid sequences, outperforming existing methods and showing utility in drug discovery.

Uni-Clip improves protein-ligand interaction modeling for undruggable and druggable targets using multimodal contrastive learning, showing significant performance gains in benchmark datasets.

Amalga improves protein backbone designability by using folding and inverse folding guidance to generate more designable structures.

Uni-RNA, a deep learning model pre-trained on a large RNA sequence corpus, is reported to lead existing methods on structural and functional prediction tasks after fine-tuning.

Uni-Fold MuSSe predicts protein complex structures from primary sequences using pre-trained protein language models, showing potential for complex prediction.

UF-Symmetry improves prediction of large protein complexes by leveraging symmetry, achieving faster computation and better accuracy than existing methods.

PyTorch reimplementation of AlphaFold and AlphaFold-Multimer with training code, retrained from scratch to equal or better accuracy, about 2.2 times faster training and roughly 2% higher multimer TM-Score.