To follow up on my second question, regarding why potential energy distribution overlap is used to judge phase-space overlap, I found this description in a 2005 paper, (J. Chem. Phys. 123, 054103 2005)
"Energy is the key quantity characterizing the importance
of a configuration to a given system, so it is appropriate to
consider energy distributions when trying to understand
phase-space relations."
which seems to explain why energy overlap is used to judge phase-space overlap. But does it hold true in practical cases? Because a big ligand and a small ligand can have very different potential energies that do not overlap, like a long alkyl-substituted benzene and a simple benzene, but the larger molecule should well include the phase space of the smaller molecule. Vice versa, if 2 ligands bind to a single large protein pocket, but occupies different ends of the pocket, they can have similiar potential energies, but have no phase-space overlap.
From what I imagine, to judge phase-space overlap, shouldn't one run independent simulations of both the starting state and final state, and see if their spatial occupancies, after reaching equilibrium state, overlap?
------------------------------------------------------------------
Sender:Liao via AMBER <
amber.ambermd.org>
Sent At:2022 Oct. 7 (Fri.) 00:30
Recipient:AMBER Mailing List <amber.ambermd.org>
Subject:[AMBER] Phase space overlap questions in alchemical free energy calculations
Dear Experts,
When doing alchemical free energy simulations, as I've read when starting state A completely overlaps final state B, this is the best scenario and simply using the phase space of A to directly sample B one can get a good free energy difference estimate (without need of intermediate states). But this is quite puzzling to me. To begin, I could not exactly think of a practical example of such a case. Would appreciate it if someone can provide a good example, where the phase space of A is much larger than B and completely overlapping B? Maybe that would alleviate my questions further down the line. Thanks so much.
Another parallel question would be that, it seems that when judging phase space overlap, people look at the potential energy distribution overlap. But that's also puzzling to me-what I imagine is that, 2 molecules can have hugely different potential energies, but sample overlapping phase space. Such example I could suppose, is A: a benzene ring with a long flexible carbon chain, and B: a simple benzene ring, both in water. Wouldn't A (alkylbenzene) cover all the phase space of B (benzene)? Or is this a wrong example of A being much larger than B in phase space and completely overlapping B. Thanks again.
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Fri Oct 07 2022 - 12:30:02 PDT
