Predicting multiple conformations via sequence clustering and AlphaFold2.

AlphaFold2 (AF2) 1 has revolutionized structural biology by accurately predicting single structures of proteins. However, a protein's biological function often depends on multiple conformational substates 2 , and disease-causing point mutations often cause population changes within these substates 3,4 . We demonstrate that clustering a multiple sequence alignment (MSA) by sequence similarity enables AF2 to sample alternate states of known metamorphic proteins with high confidence. Using this method, AF-Cluster, we investigated the evolutionary distribution of predicted structures for the metamorphic protein KaiB 5 , and found that predictions of both conformations were distributed in clusters across the KaiB-family. We used nuclear magnetic resonance (NMR) spectroscopy to confirm a surprising AF-Cluster prediction: a cyanobacteria KaiB variant is stabilized in the opposite state than the more widely-studied variant. To test AF-Cluster's sensitivity to point mutations, we designed and experimentally verified a set of 3 mutations predicted to flip KaiB from Rhodobacter sphaeroides from the ground to the fold-switched state. Finally, screening for alternate states in protein families without known fold-switching identified a putative alternate state for the oxidoreductase Mpt53 in M. tuberculosis. Further development of such bioinformatic methods in tandem with experiments will likely have profound impact on predicting protein energy landscapes, essential for illuminating biological function.