RBFE of protein mutations

[ABarancewicz]

Dear OpenFE devs and users,

I'm interested in predicting effects of protein mutations on affinity to a ligand. I have limited experience with MD and FEP, so I'm unsure what challenges are associated with such a simulation, although I believe this has been done in previous studies (https://doi.org/10.1016/j.jmb.2019.02.003). I have also noticed that this has been mentioned in a few issues (#682, #633, #634, #636), and have pieced together that it may be possible although there is no easy way to make protein-protein mappings currently.

Does OpenFE allow for alchemical transformations of the protein to investigate the effect of mutations? Would working with GROMACS be more suitable for this? Any insights on the best way to conduct such a simulation would be very much appreciated!

Thanks,
Anthony Barancewicz

[IAlibay]

Hi @ABarancewicz,

Protein mutations are not currently available in OpenFE, however @ijpulidos is currently working towards implementing them. In the first instance it will be for NECycling in feflow. I'll let him weigh in with more details.

[ijpulidos]

@ABarancewicz It's definitely possible to do with the OpenFE stack. For the protein-protein mappings, kartograf is already capable of making these.

I've been working on a protein-mutation-capable protocol that I'm hoping we'll finally get released soon in the following weeks. If you want more details about it, I'm working on it in https://github.com/OpenFreeEnergy/feflow . Happy to answer more questions that you may have.

[ABarancewicz]

Thanks so muich @IAlibay and @ijpulidos, really cool work that you guys are doing! I'll take a look at kartogragf and keep an eye out for the feflow release.