Dear gmx_MMPBSA users,
My system of interest is a homotetrameric protein in which each subunit can bind a ligand. I seem to have successfully calculated the binding energy of one ligand with the tetramer using gmx_MMPBSA. In this case, the ligand interacts mainly with one of the monomers. I have two questions about that:
1. Suppose I have a trajectory in which each subunit is bound to a ligand. Is it theoretically possible and or/appropriate to calculate the binding energy (e.g., GB + IE) for each ligand? I suppose it would be computationally convenient to do so because then I could increase the number of replicates and thus save on computational resources.
2. If the answer to 1 is positive, then how can I account for the presence of the other ligands when iteratively calculating the binding energy of each ligand? Can I simply use an index file with a receptor group containing the tetramer and the other three ligands?
3. Still if 1 is positive, can gmx_MMPBSA handle this specific problem automatically or at least automate the work at any step?
Thank you in advance.
Best regards,
Gustavo