End-to-end differentiable learning of protein structure

It would be cool if machine learning researchers would start participating CASP and CAPRI. If you crack Go, you get fame, but if you crack protein prediction, you get Nobel price and completely revolutionize biochemistry and medicine.<p><a href="http:&#x2F;&#x2F;predictioncenter.org&#x2F;" rel="nofollow">http:&#x2F;&#x2F;predictioncenter.org&#x2F;</a><p><a href="http:&#x2F;&#x2F;www.ebi.ac.uk&#x2F;msd-srv&#x2F;capri&#x2F;" rel="nofollow">http:&#x2F;&#x2F;www.ebi.ac.uk&#x2F;msd-srv&#x2F;capri&#x2F;</a><p>edit: Why there is no XPRICE for protein folding?

I work on protein structure, albeit not from a computational standpoint, and it struck me as odd that none of the work from the Baker group (Univ Washington) e.g. Rosetta (<a href="https:&#x2F;&#x2F;www.rosettacommons.org&#x2F;" rel="nofollow">https:&#x2F;&#x2F;www.rosettacommons.org&#x2F;</a>) was mentioned. Rosetta can be used to predict tertiary structure from amino acid sequence. Does anyone familiar with the field know how the methods used by software like ROSETTA differ from those presented in this paper?

This is a very interesting approach. Clearly a lot more work to do but the robust prediction of protein structure from sequence would be an absolute game changer for biomedical science so I hope that this opens up new strategies.

What are the real-world applications of protein folding (preferably, some specific example)? I always hear that it&#x27;s really important for drug design and biotechnology but have a hard time imagining something concrete.

Cool method! Are you planning to participate in the next CASP? Do you plan to open source the code?