Re: Problem with iron ion (FE2+)

From: Tristan Croll (
Date: Tue Aug 25 2015 - 03:56:51 CDT

In general, the parameterisation of multivalent metal ions in CHARMM still seems quite poor - the strength of interactions with negatively charged groups is far too strong. It's an interesting exercise to play around with a magnesium-ATP complex in implicit solvent - you'll find you can get it to stably contort into all sorts of "interesting" arrangements.

From: <> on behalf of Josh Vermaas <>
Sent: Tuesday, 25 August 2015 5:17 AM
To:; Branko; Daniela Rivas
Subject: Re: namd-l: Problem with iron ion (FE2+)

Hi Daniela,

I've usually seen metal ions inside proteins treated as being covalently linked to their natural protein interaction partners. Either it is done explicitly in building up the topology (heme is the classic example of this), or if no other parameters are available, NAMD extrabonds are used to mimic the interactions the ion would have with its surroundings (usually with force constants and equilibrium lengths in line with what was seen in the crystal structure). Harmonic restraints are sometimes unsuitable, especially if you expect the protein to move, as the harmonic restraints won't move with a diffusing protein, causing the ion to drift within the protein.

-Josh Vermaas

On 08/24/2015 01:40 PM, Branko wrote:

I suppose that your ion is not bounded to protein, but you initially asked how to preclude movement of this ion. - probably toward solvent. Try to read what kind of restraints NAMD offer and to conclude which one is most suitable for your system. I suppose that harmonic restraint or extra bond will be most suitable. Also, if use harmonic restraint, in few separate short simulations you can find most suitable 'force value' with which you restraint ion


On 8/24/2015 8:16 PM, Daniela Rivas wrote:
Hi Branko,

What kind of restraint?
Also, I forgot to mention that the iron ion is not bonded to the protein, it's just interacting with the surrounding residues. So, I want it to move during the simulation, that's what I want to look at.


2015-08-24 15:09 GMT-03:00 Branko <<>>:

If you 100% confined that Fe2+ ion should be in position in which you introduced it initially, try to impose restraint to this ion.>

Also impose restraints to all counter-ions that you have in system under consideration, especilay to negatively charged ones.


On 8/24/2015 7:51 PM, Daniela Rivas wrote:

I'm trying to run a molecular dynamics with Namd with an introduced iron ion (FE2+) inside the protein, where is suppose to be. I checked my ionized PDB and PSF and everything was perfect, but when the simulation started I found that the iron was not in the position where it was supposed to be and it was touching another atom, which is bad because I can't analyse that.

Any thoughts about what's happening? I've done every step twice and I get the same result.

Thanks in advance,

This archive was generated by hypermail 2.1.6 : Tue Dec 27 2016 - 23:21:16 CST