AMBER Archive (2009)

Subject: Re: [AMBER] Heme + Fe(II) + O2 + histidine parameter

From: Haining Liu (
Date: Thu Sep 03 2009 - 16:13:09 CDT

Dear Francois,

My next question is that how should I set the VdW radii for Fe in
Gaussian calculation. Is it 1.8 as default in RED?


FyD wrote:
> Dear Haining Lu,
>> I want to run a MD simulation on hemoglobin. My question is how to get
>> the parameter for heme? From a literature search, people seem to
>> consider the (heme + O2 + Fe(II) + histidine) as a whole residue and
>> calculate the RESP charge using the Gaussian program.
> I think you should condider all the ligands connected to the Fe(II) &
> not only the porphyrin ring...
>> However, if such
>> a system is used in Gaussian, the histidine residue is not complete,
>> i.e., its backbone is cut from crystal structure. In that case, how can
>> I determine the total charge of the system in order to run the quantum
>> calculation? The second question is how to let the AMBER program know
>> that the histdine should be included in heme? Do I have to modify the
>> PDB file?
> Here, the idea is to generate a molecular fragment (i.e. a new force
> field library) for your heme system. Your case is a little complex but
> you will find many examples of generation of basic molecular fragments
> @ The presence of the metal
> atom makes the system more complex (mainly in QM & during fitting
> step) but does not change the strategy for building a molecular
> fragment. You build such a molecular fragment from a whole molecule
> from which some atoms are removed. The group of atoms which is removed
> is usually constrained during the charge fit (using the RESP program).
> At the end, the fragment built has to be compatible with the other FF
> libraries (belonging to the FF you chose) used to construct/recognize
> the entire biopolymer (containing the metal complex).
> Finally, to answer to "how to let the AMBER program know that the
> histdine", the LEaP program is used for automatic recognition between
> the PDB file (biopolymer with the metal complex) and the FF libraries
> required: this recognition is simply based on residue & atom names. If
> a perfect match between the residue & atom names found in the PDB file
> & these in the FF libraries previously loaded occurs, you will be able
> generate the prmtop/prmcrd files for MD simulation.
> These are only generalities & will not help that much.
> May be, you could first read the tutorials @
> and then come back with more specific questions regarding how to built
> your metal complex fragment.
> regards, Francois
> _______________________________________________
> AMBER mailing list

AMBER mailing list