AlphaFold knows very little about the physicochemical mechanisms governing protein folding and doesn’t give any new insights either. The way it samples conforma

AlphaFold knows very little about the physicochemical mechanisms governing protein folding and doesn’t give any new insights either. The way it samples conformational space is very different from Folding@home’s molecular dynamics based approach, for instance. To describe the folding of a protein chain, a “picture” of its lowest energy conformation (which is the goal of AlphaFold—and any structural modelling tool) is not enough. We rather need a “movie” showing the steps said chain undergoes, from an unfolded polypeptide to its final, folded, native state (the “picture”).