OpenFold: retraining AlphaFold2 yields new insights into its learning mechanisms and capacity for generalization.

AlphaFold2 revolutionized structural biology with the ability to predict protein structures with exceptionally high accuracy. Its implementation, however, lacks the code and data required to train new models. These are necessary to (1) tackle new tasks, like protein-ligand complex structure prediction, (2) investigate the process by which the model learns and (3) assess the model's capacity to generalize to unseen regions of fold space. Here we report OpenFold, a fast, memory efficient and trainable implementation of AlphaFold2. We train OpenFold from scratch, matching the accuracy of AlphaFold2. Having established parity, we find that OpenFold is remarkably robust at generalizing even when the size and diversity of its training set is deliberately limited, including near-complete elisions of classes of secondary structure elements. By analyzing intermediate structures produced during training, we also gain insights into the hierarchical manner in which OpenFold learns to fold. In sum, our studies demonstrate the power and utility of OpenFold, which we believe will prove to be a crucial resource for the protein modeling community.

Single-sequence protein structure prediction using a language model and deep learning.

AlphaFold2 and related computational systems predict protein structure using deep learning and co-evolutionary relationships encoded in multiple sequence alignments (MSAs). Despite high prediction accuracy achieved by these systems, challenges remain in (1) prediction of orphan and rapidly evolving proteins for which an MSA cannot be generated; (2) rapid exploration of designed structures; and (3) understanding the rules governing spontaneous polypeptide folding in solution. Here we report development of an end-to-end differentiable recurrent geometric network (RGN) that uses a protein language model (AminoBERT) to learn latent structural information from unaligned proteins. A linked geometric module compactly represents Cα backbone geometry in a translationally and rotationally invariant way. On average, RGN2 outperforms AlphaFold2 and RoseTTAFold on orphan proteins and classes of designed proteins while achieving up to a 106-fold reduction in compute time. These findings demonstrate the practical and theoretical strengths of protein language models relative to MSAs in structure prediction.