MD simulations of coronavirus protein ORF3a - ligand complexes

Necessary run scripts and parameters to run MD simulations in publication (preprint version) https://www.biorxiv.org/content/10.1101/2025.03.20.644376v1 The two scripts job_ligand_* contain the entire MD simulation workflow for a Protein Ligand complex using parametrization with Espaloma in OpenMM (done with the Python script https://github.com/imec-int/md_sarscov2_orf3a/blob/main/openmm2gmx_protein_ligand.py) and running multiple energy minimization steps followed by equilibration, pulling, and the actual MD simulation run. The output of the pulling simulation is used as a starting point for running umbrella samplings of the protein-ligand complex. This workflow is contained at the bottom of the scripts as a series of function calls in bash:

parametrize_system ${PARAMETRIZATION}
prep_vacuum ${OUT_NAME}_${PARAMETRIZATION}
min_0_vac ${OUT_NAME}_${PARAMETRIZATION}
min_1_vac ${OUT_NAME}_${PARAMETRIZATION}
min_2_vac ${OUT_NAME}_${PARAMETRIZATION}
prep_solvated_from_vacuum ${OUT_NAME}_${PARAMETRIZATION}
min_0 ${OUT_NAME}_${PARAMETRIZATION}
min_1 ${OUT_NAME}_${PARAMETRIZATION}
min_2 ${OUT_NAME}_${PARAMETRIZATION}
min_3 ${OUT_NAME}_${PARAMETRIZATION}
min_4 ${OUT_NAME}_${PARAMETRIZATION}
npt_5 ${OUT_NAME}_${PARAMETRIZATION}
npt_6 ${OUT_NAME}_${PARAMETRIZATION}
pull_after_npt6 ${OUT_NAME}_${PARAMETRIZATION}
npt_7_prep ${OUT_NAME}_${PARAMETRIZATION}
npt_7 ${OUT_NAME}_${PARAMETRIZATION}

Going through this in order: The first call parametrizes the system, for which a PDB file of the complex and an SDF file of the ligand with coordinates relating to the binding pocket is required. Then the system is put into a simulation box with periodic boundary conditions, but no solvent yet. The first three minimization steps are executed to relax the complex in vacuum. Then with prep_solvated_from_vacuum the equilibration and charge neutralization are done, followed by 5 more minimization steps. Followed by preparations and a final MD run in the NPT ensemble. This is based on the protocol and series of .mdp files described here https://github.com/carlocamilloni/labtools/tree/main/mdps/atomistic

In pull_after_npt6 the necessary COM pulling between ligand and protein is performed. This generates the necessary trajectory to pick windows from for the umbrella sampling run. To perform the umbrella sampling, the Jupyter notebooks https://github.com/imec-int/md_sarscov2_orf3a/blob/main/PreUmbrellaAnalysis.ipynb and https://github.com/imec-int/md_sarscov2_orf3a/blob/main/PostUmbrellaAnalysis.ipynb are used together with the generic job scripts for running umbrella windows: https://github.com/imec-int/md_sarscov2_orf3a/blob/main/job_ligand_bacteriochlorin and https://github.com/imec-int/md_sarscov2_orf3a/blob/main/job_ligand_bacteriochlorin_umbrella .

Finally, in https://github.com/imec-int/md_sarscov2_orf3a/blob/main/run_mmbsa_SB431542 an example of how to run MMPBSA with gmx_MMPBSA can be found, using the parameters defined in "mmpbsa.in".

Ligands

The two used ligands here are SB431542 and a bacteriocholorin-like molecule.