Does AlphaFold predict the spatial structure of a protein from physics or recognize it (its main parts and their association) using databases? https://www.biorxiv.org/content/10.1101/2022.11.21.517308v1?med=mas
@albertcardona @biorxivpreprint the preprint doesn't show anything of what it is claimed in the abstract.
@pedrobeltrao @biorxivpreprint The are some issues. Concerning are the 4x private communications from "J. Jumper", and that the conclusion that the success is from "recognition based on similarity" seems to derive not so much from the statistical model in the results but from a point made in the conclusions section on the number of parameters used by AlphaFold 2.
@biorxivpreprint “AlphaFold recognizes a 3D structure of the examined amino acid sequence by a similarity of this sequence (or its parts) to related sequences with already known 3D structures. Concluding, I have to emphasize that this paper does not diminish the merit and utility of AlphaFold; it only explains the basis of its success.”
Had a good laugh. Like revealing the Wizard of Oz but recognizing the power of its perfomance.