Re: detecting vanishing atoms in zero.fep

From: Brian Radak (brian.radak_at_gmail.com)
Date: Thu Nov 01 2018 - 16:59:58 CDT

Looking at your files, I'm seeing way more issues than just the alchemical
flags. How do you have multiple atoms in the same molecule with the same
name? The names also do not at all appear to match the residue in your RTF
from the server. I might suggest constructing the molecule in a
non-alchemical PSF first and making sure that you can simulate it in NAMD.
Then, produce a new residue or alchemical patch (PRES) that will create the
alchemical system. This should also make it easier to keep track of which
atoms belong in each group.

As per your other comments, it sounds like you are using only one set of
alchemical flags at a time - you need to define both groups in the ".fep"
file so that interactions between the endpoints can be noted and excluded.
Not doing this almost always leads to large forces and unstable dynamics.

HTH,
BKR

On Thu, Nov 1, 2018 at 4:40 PM soroush ziaei <soroush.ziaei7337_at_gmail.com>
wrote:

> i have to calculate solvation free energy ( third part of tutorial ) in
> order to achieve delta<g> for compare the solvation between two structure.
> I have showed topology file recently (generated with cgenff & swiss param
> server). this structure has a symmtry similar to hybrid molecule in
> tutorial file & i want to know how can i detect which atoms should be -1,1
> or 0 ?!
> my prespective ( according to the fep file that already exist in tutorial
> with symmetry in some groups ) is:
> first, i have to change the value of one side to 1 and the up side
> (similar to tutorial) to -1, doesn't it?
> if my guess is almost correct, then why after changing value of one side
> to 1, the running of fep for backward faced to an error? [rm, can not
> remove eq3.restart; no such file or directory]!
>
> i appreciate,
> please guide me.
>

This archive was generated by hypermail 2.1.6 : Mon Dec 31 2018 - 23:21:29 CST