Re: [AMBER] Regarding Protein-ligand MD

From: Bikash Ranjan Sahoo <>
Date: Tue, 28 Oct 2014 12:17:02 +0900

Dear Dr. Robin,
      I am glad to have your prompt response. I am very new to Amber. This
approach I had made for solving the problem using Gromacs. I just have
sliced the ligand and attached residues, and generated the topology using
PRODRG. I fortunately got the bonded information, however failed to ran MD
due to numbering and other inconsistency. If here also I will do the same,
how can I grep the informations to the protein topology. The problem is
renumbering all atom numbers, arranging all residues according to original
protein numbers etc. I understand from the tutorial is that it will take
long time for calculation. Thank you very much for letting me know about
the computational inconsistency for ~200 AAs. Can you please suggest me how
to combine the parameters? I mean, after generating topology the bonded
informations will be i-1, i, i+1 (where "i" is ligand). Now when I will
combine this with protein, simply I have to silent i-1 and i+1 parameters.
But, the problem here is the program does not recognize current "i" atom
numbers. When I renumbered them using my scripts and modified the bonded
information by manually calculation, I found the error in atomic number
inconsistency. Even, I tried to generate the itp file for ligand only and
ligand with i-1 and i+1. Then I just provided the covalent informations
(bond, angle, dihedrals etc). Still I am unable to fix the problem. So
kindly suggest me if I will generate it using Param fit, how can I combine
the system with proper atomic declaration?

Thank you very much Dr. Robin.


On Tue, Oct 28, 2014 at 12:02 PM, Robin Betz <> wrote:

> Hi Bikash,
> Creating parameters not present in conventional force fields is what
> Paramfit was designed for.
> However, part of the process involves doing QM calculations on many
> conformations of your system, and 200 AA is much too large for this. I
> would suggest taking a subset of your structure, for example just the
> ligand and whatever amino acids it is bound to, and fitting parameters to
> that smaller system.
> Best,
> Robin Betz
> On Mon, Oct 27, 2014 at 7:58 PM, Bikash Ranjan Sahoo <
>> wrote:
> > Dear All,
> > I am facing some problems in simulating a P-L complex where
> the
> > ligand is covalently bonded to the protein with two adjacent residues.
> The
> > ligand is a unnatural amino acid. In such scenario, I came across the
> > Paramfit tutorial where it is possible to generate a FF from the inputs.
> Is
> > it possible to generate a force field for such a system that will hold
> all
> > connecting information (covalent bonded information) between the ligand
> > and protein. If it is possible, can somebody suggest how long it will
> take
> > to generate force field parameters for ~200 AA long system.
> >
> >
> > Best Regards
> > Bikash
> > _______________________________________________
> > AMBER mailing list
> >
> >
> >
> _______________________________________________
> AMBER mailing list
AMBER mailing list
Received on Mon Oct 27 2014 - 20:30:04 PDT
Custom Search