On Thu, Aug 29, 2013 at 5:00 PM, Fabrício Bracht <fabracht1.gmail.com>wrote:
> Hello. I want to calculate the pka of two residues and I have decided to
> use the constant pH simulations in amber to calculate the titration curves
> of these residues. I am running replica exchange with constant pH and I am
> trying to follow the "tutorial" given in
> http://jswails.wikidot.com/ph-replica-exchange#toc4
> I 've noticed a few disparities between what is shown on that wiki and what
> really happens with my calculations. A few discrepancies include:
> The ntxo flag is set to 2 in the mdin file for the replica exchange
> calculations, but I found no such value on the manual.
>
See
http://ambermd.org/doc12/ regarding the ntxo variable.
> The page says "When running replica exchange simulations, the format of the
> cpout file is slightly changed so that the current target pH value is
> printed next to each record". But this is not what has happened here. My
> cpout files contain:
>
> Solvent pH: 3.50000
> Monte Carlo step size: 5
> Time step: 5
> Time: 1000.008
> Residue 0 State: 0
> Residue 1 State: 0
>
> Residue 1 State: 0
>
> Residue 1 State: 0
> .
> ..
> But no pH value is printed next to each record.
> I have put a hold on the simulations in order to check if pH-REMD
> simulations are still a working feature in amber12 and if this is the
> behaviour I should expect from the simulations.
>
Alas, one of the downsides of storing so much stuff on my Wiki. I had
originally written that page to show collaborators how to use my pH-REMD
code, but it is only available in development versions of Amber at the
moment (I added it in the summer of 2012, a couple months after Amber 12
was released). With the aim of making that tutorial 'future-proof', or
'future-resistant' more likely, I added a statement that Amber 14 or later
(due to be released in April) is required to run pH-REMD near the top of
that page, since it will definitely be available then (on GPUs, even! -- it
is blazing fast).
That tutorial is fully up-to-date with regards to simulation output (as I
just updated it a few days ago).
All the best,
Jason
--
Jason M. Swails
BioMaPS,
Rutgers University
Postdoctoral Researcher
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Thu Aug 29 2013 - 15:00:03 PDT
