From: Paweł Kędzierski (
Date: Tue Apr 23 2013 - 03:13:08 CDT

I am not familiar with heme modeling but this may be a useful hint.
Angle and dihedral parameters are not in general sufficient to enforce
planarity. For this purpose one needs "out-of-plane" harmonic force
constants. These are called "improper dihedrals" in charmm forcefield or
simply impropers, as the independent variable is a dihedral angle
defined by four atoms, but the energy term is harmonic instead of periodic.
It seem possible to me that at creation of your own heme topology file
you may have omitted the definition of these planarity enforcing parameters.
Please consult the online charmm documentation, e.g., but in short you
must provide IMPR lines in the residue topology definition (four atom
names defining the dihedral, central atom first), and of course the
relevant parameters must be present in the IMPROPER section of the
parameter file. I am sure these parameters are available for heme so it
may only be an omission in the residue topology definition.

W dniu 22.04.2013 21:10, Yarrow Madrona pisze:
> Hi Axel,
> Thank you for your help. I have been learning MD for only a month now so
> please bare with me. I apologize in advance for this long winded post.
> Unfortunately, I do not know anyone who has done extensive simulations
> with cysteine HEME linkage using CHARM parameters. The few I see on line
> look similar to what I have done but I have no way of knowing whether they
> see flex in the heme or not.
> 1. To answer your first question: The first simulations were done with the
> top_par_22_heme.str streaming file that is input into NAMD. I had the same
> problems with some non-planarity. kind of looks like a v with the iron at
> the apex (about 150 to 160 deg). There was also a distance contrain
> between the cysteine and iron.
> I have heard in a forum that for a pentacoordinate heme like Myoglobin
> with Histine coordinating it is better to use the PHEM; His bound to Fe
> patch in the heme streaming file. It was mentioned that the "The FHEM
> patch is needed to remove a couple N-Fe-N ANGLE terms added by
> I was having a little difficulty understanding how to implement the
> patches so I built the parmameter file/topology file for a HEME + Cystene
> residue I called HEC. I defined bonds, Improppers and internal
> coordinates in the topology file. Most of these come straight from the
> HEME streaming file with corrections for the new cysteine bond. I also
> entered the spring constant, distance and angle in the parameter file.
> Everything behaves well except for that that flexing of the Heme.
> The only thing that is not in there from the heme streaming file is the
> following:
> ANGLE 1C1 1S 2FE 1C1 1S 2FE
> ANGLE 1S 2FE 2NA 1S 2FE 2NB 1S 2FE 2NC 1S 2FE 2ND
> IC 1C1 1S 2FE 2NA 0.0000 0.0000 0.0000 0.0000 0.0000
> IC 1C1 1S 2FE 2NB 0.0000 0.0000 0.0000 0.0000 0.0000
> IC 1C1 1S 2FE 2NB 0.0000 0.0000 0.0000 0.0000 0.0000
> IC 1C1 1S 2FE 2NB 0.0000 0.0000 0.0000 0.0000 0.0000
> I added the IC's to the parameter file but it does not seem to make a
> difference. I suppose I can add the delete comment into the psfgen before
> running it. Should I manually put in the length, dihedral and angle of
> all of the above?
The IC lines are only used by psfgen if some atoms are missing in PDB
coordinates; internal coordinates are then necessary to generate their
positions relative to present ones. These lines are not used for
parametrization so most likely you don't need them at all.

> I already have defined may of them in addition to the ones taken from Heme
> streaming file:
> IC -C CA *N HN 1.3479 123.9300 180.0000 114.7700 0.9982
> IC -C N CA C 1.3479 123.9300 180.0000 105.8900 1.5202
> IC N CA C +N 1.4533 105.8900 180.0000 118.3000 1.3498
> IC +N CA *C O 1.3498 118.3000 180.0000 120.5900 1.2306
> IC CA C +N +CA 1.5202 118.3000 180.0000 124.5000 1.4548
> IC N C *CA CB 1.4533 105.8900 121.7900 111.9800 1.5584
> IC N C *CA HAY 1.4533 105.8900 -116.3400 107.7100 1.0837
> IC N CA CB SG 1.4533 111.5600 180.0000 113.8700 1.8359
> IC SG CA *CB HB1 1.8359 113.8700 119.9100 107.2400 1.1134
> IC SG CA *CB HB2 1.8359 113.8700 -125.3200 109.8200 1.1124
> IC CA CB SG FE 1.5584 113.8700 180.0000 111.5140 2.2000
> Thank you.
> -Yarrow
>> two comments:
>> 1) have you made some test calculations with the regular heme residue as
> provided in the parameter set to have a reference to compare to? 2)
> parts of molecules are usually held flat through improper
>> dihedrals. have you checked whether they are exist where needed and have
> assigned parameters for your custom residue? again, it can be extremely
> helpful to compare to "vanilla" reference.
>> axel.
>> On Mon, Apr 22, 2013 at 12:49 AM, Yarrow Madrona <>
> wrote:
>>> Hello,
>>> I know this is a CHARMM question but I am having trouble finding
> information on this.
>>> I am wondering if anyone here has experience with HEME-cystiene Fe
> thiol
>>> or Fe-HIS simulation. I recently build a parameter and top file with a new
>>> residue "HEC" which is a combination of the HEME and cysteine as a single
>>> residue. I took most of the HEME parameters from the CHARMM Heme streaming
>>> toppar file.
>>> Everything looks fine except the HEME seems to flex a little. Looks a
> little like a "V" Maybe 140-160 deg instead of 180 degrees across the
> heme. I have heard that there are some parameters to take from the PRES
> SUL patch in the HEME topar streaming file. I have taken and modified
> the
>>> relevant ones but I was wondering if anyone knows specifically, what would
>>> cause the heme to be less than ideal planar.
>>> Thank you for your help.
>>> --
>>> Yarrow Madrona
>>> Graduate Student
>>> Molecular Biology and Biochemistry Dept.
>>> University of California, Irvine
>>> Natural Sciences I, Rm 2403
>>> Irvine, CA 92697
>> --
>> Dr. Axel Kohlmeyer
>> International Centre for Theoretical Physics, Trieste. Italy.