You cannot compare them that way, so what about a swap. Create a tripeptide with the covalently attached ligand that is distant from the protein and then swap the co-factor states in two separate simulations (one w/ charge on protein, one with charge on peptide). Still not guaranteeing the energy differences will be meaningful, but the prmtops will have equivalent presumably… Alternatively leave the net-negative charge; that will not break the PME or try both approaches to demonstrate they get similar trends.
-tec3
> Dear Thomas and all,
> In my case, between the two redox-states, there is a charge difference in
> the cofactor-covalently attached to the protein.
> And because I'm neutralizing the system, tleap adds one Cl- anion extra
> when the co-factor is more positive compared to the other redox-state.
>
> So my question is, is it meaningful to simulate a system with non-zero
> net-charge (+1 in this case) and compare the energies with a neutral system
> (which has an equal number of protein, water, ion and co-factor atoms)? If
> yes, I'm thinking of generating a matrix of potential energy differences
> (printed in mdout files) between all the snapshots. Will the generation of
> multiple trajectories for both systems and then re-estimating this E-matrix
> be able to deal with the noise?
>
> I'm reading your papers carefully with the hope that the answers might be
> already there.
> Nevertheless, I rephrased my question(s) here with the hope that other
> experts on the list might also give valuable suggestions.
>
> Thanks a lot for valuable comments/suggestions.
> Best regards.
>
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
- application/pkcs7-signature attachment: smime.p7s
Received on Thu Oct 08 2020 - 17:00:02 PDT
