From: L. Michel Espinoza-Fonseca (
Date: Tue Nov 03 2009 - 05:24:46 CST

Hi Luis,

There are two things that you can do to solve this issue. Each solution
depends on the residue you are interested in.

1. For histidine, you can easily select the residues and change their names
with VMD, so that psfgen can identify the "right" protonation state. For
instance, if propka tells you that histidines X and Y have their Nepsilon
protonated and Y is fully protonated, you are going to rename them as HSE
(for X and Y) and HSP (for Z). You can achieve this by using VMD:

set sel [atomselect top "resid Y"]
$sel resname HSP

Then you just use psfgen and the program should be able to recognize the new
names and assign the correct protonation state.

2. As for protonated Asp and Glu, you will need to apply the corresponding
patch (ASPP and GLUP, respectively).

If you do everything correctly, solvation and ionization should not have any
effect on the protonation states you already assigned with psfgen.


On Tue, Nov 3, 2009 at 11:50 AM, Luis Agullo (LAB) <> wrote:

> Hello,
> I use the functions of autopsfgen, solvation and ionization to prepare
> proteins for NAMD. However, I would like to change the charge of specific
> amino acids in my protein as estimated by other programs (Propka,...). I
> tried to do it with Molefacture, deleting or adding protons to the
> structure (the output of autopsfgen); but, after using the other functions
> (solvation, ionization), VMD delete those changes and I obtain finally the
> initial ionization pattern of my protein. How can I solve that?
> Luis