SIRAH LLPS of FUS-LC

[gosthak]

Dear SIRAH Developers,

I recently attempted to use SIRAH-ff to investigate whether it could simulate the phase separation of intrinsically disordered proteins (IDPs), specifically the well-studied FUS-LC. I was motivated by the availability of explicit solvent modeling and the ability to run simulations on a GPU.

For this experiment, I referred to your comprehensive tutorial at https://sirahff.github.io/documentation/Tutorials%20gromacs.html#proteins-in-explicit-solvent, with a few modifications:

I placed 40 molecules of FUS-LC into a large box with a minimal amount of water to implement NPT and induce condensation. I then proceeded with the energy minimization and equilibration steps as outlined in your tutorial, extending the duration by approximately five times to accommodate the larger system size.

I did not impose any restraints, considering the inherently unstable structure of IDPs and the need for water molecules to move freely, facilitating phase separation of the protein condensate.

Despite these measures, I observed no phase separation. I even employed an annealing scheme in the md.mdp to test various temperatures, extending to non-realistic levels above 400K, to determine if the issue was related to energy scale. Interestingly, the protein condensate became more compact as the temperature increased, which is atypical.

I have attached the *.gro snapshots (the filenames include the time in picoseconds) and the run-mdp file for your review.

Could you please provide any suggestions on optimizing the configuration for liquid-liquid phase separation of IDPs? If further details are needed, I am happy to provide them.

Sincerely,
Gosha
sirah_fus-lc.tar.gz