I want to predict the effect of a point mutation in a protein stability. I have the .pdb crystal structure of the wild type.
After a literature search, I think this paper presents the most comprehensive study of different methods for this analysis DOI:10.1002/prot.22921. So I want to do the prediction using their protocol 16. I also found python code for this protocol here: https://github.com/Kortemme-Lab/ddg/tree/master/protocols/ddg_monomer_16.
My question is if anyone has setup a Rosetta Scripts .xml file to reproduce this protocol?
Note that the Kortemme lab scripts you point to aren't actually running the protocol in PyRosetta, instead they're job-handling scripts which invoke (though a subshell) the actual ddg_monomer application.
Unfortunately, the mover which forms the core of the Kellogg ddg_monomer protocol is not accessible through RosettaScripts. That said, you should be able to be able to put together a sequence of PackRotamersMovers and MinMovers which approximate the protocol being used by the ddg_monomer application.
The tricky bit would be the constraints which are added by the ddg_monomer application. For minimization, these are Calpha-Calpha distance constraints for all protein residues within 9 Ang. I belive that a close approximation can be obtained with the AddConstraintsToCurrentConformationMover, changing `use_distance_cst` to true, `max_distance` to 9.0 and coord_dev to 0.5.
Since the Kellogg et al paper, have there been improvements in using ROSETTA for prediction of stability ddG's of mutations in single proteins? Is there an updated protocol?