Dear Ambers,
I wonder if there is a possibility in AMBER at the moment to calculate
the enthalpic effect of explicit solvent to binding. Recently I
encountered this paper:
http://pubs.acs.org/doi/full/10.1021/jacs.5b08717
where, if I correctly understand the authors, they extracted internal
energy corresponding to solute-solute, solute-solvent and
solvent-solvent interactions from the trajectory using H=E+pV and
ignoring pV contribution as negligible. They used CHARMM. I wonder,
would it make sense to do something like this in AMBER:
- to extract solute-solute interactions from the trajectory of the
complex using LIE or simply in vacuo ELE+VDW from MM-PBSA/MM-GBSA;
- for three long enough separate trajectories of complex, receptor and
ligand in explicit solvent to define always the same amount of the water
molecules around the solutes and extract the corresponding trajectories
using 'closest' command, and then to use min=5 to estimate all ELE and
VDW contributions corresponding to non-bonded interactions in the system
(solvate-solvent and solvent-solvent).
Because the number of water molecules for each system (complex,
receptor, ligand) will be the same, the obtained numbers should be
comparable though there are some problems I would expect:
- if there are too many water molecules taken into account with the
'closest' approach, the results will be very noisy;
- if there are not enough water molecules taken into account the
differences in shape/geometry of the receptor and ligand could be
crucial for the energies obtained like this because the hydration shells
around a global protein and, for example, an elongated peptidic ligand,
have very different shapes of the volumes and, therefore, amounts of
interactions.
What is your opinion: can this kind of approach yield anything valuable
or it should be expected to be extremely noisy/unreliable anyway?
As an alternative I see MM-PBSA/MM-GBSA approaches to calculate the
enthalpy, which can be combined with umbrella sampling/SMD free energy
calculations for the full free energy G in explicit solvent. In this
case, the solvent is taken into account explicitly for G but implicitly
for H. Depending on the GBSA-model the difference G-H and, therefore,
the entropy can differ a lot quantitatively. However, if this is applied
for very similar systems, do you think a certain qualitative pattern of
the contributions of S and H components to G can be discussed at all
also in terms of solvent impact?
I found many discussions on the Mailing List from several years ago
touching this topic. However, for me it is still not clear if some
calculations taken into account solvent can be feasible or not at the
moment.
Thank you very much in advance for sharing your opinion.
Cheers,
Sergey
--
Sergey A. Samsonov
Postdoctoral researcher
Structural Bioinformatics
Biotechnology Center
Tatzberg 47-51
01307 Dresden, Germany
Tel: (+49) 351 463 400 83
Fax: (+49) 351 463 402 87
E-mail: sergey.samsonov.biotec.tu-dresden.de
Webpage: www.biotec.tu-dresden.de
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Mon Jan 18 2016 - 07:30:05 PST
