[gmx-users] Re: How to ignore atoms during simulations??

Cesar Araujo cesar.araujo at oulu.fi
Wed Jun 21 17:09:10 CEST 2006

Thanks PB, You are right to ensure invisibility also force-field parameters must
be 0.0. I think that I can do it replacing the type of those atoms that I want
to make invisible for a dummy one (e.g. IW type for Gromacs force-field is
defined as a dummy particle with all the parameters in 0.0). The idea behind
this is to get rid of LINCS warnings that are so common when you are trying to
perform a docking experiment with Gromacs. If you have to setup your ligand
near the active site that you are exploring using some molecular modelling
tool, it's very common to get LINCS warnings for residues clashing with your
ligand. Some experts in the field have suggested that one alternative is grow
the ligand near the active site. It means, start with an small portion of the
ligand, run energy minimization followed by position restraint, and then add
some another part to your ligand and repeat those steps until you have all the
ligand builded near the active site. After that, you will be able to run your
MD step without LINCS problems. One method to do this is to build those
fragments and after each cycle (EM+PR) replace the old fragment of your ligand
by the new one. But this is a very time consuming task and still requires
manual work. So, one alternative that I'm trying to test with this
"invisibility" feature, is start with the whole ligand but making visible just
few more atoms after each (EM+PR) cycle. The advantage of this, is that it
could be easily automated, because you just need to touch topology text files.
So,let's see if I can reach to some interesting result.

Any suggestion, will be appreciated :)


> I was wondering whether one can have atom types with zero charge, zero 
> vdw (C6=0, C12=0; Or sigma=0; epsilon=0), but with mass. Won't they be 
> invisible in the energy/force calculations (if they are not connected 
> by any bonded interaction)?

> pb.

On Jun 21, 2006, at 9:57 AM, Cesar Araujo wrote:

> Cesar Araujo wrote:
>> Hi,
>> Could anybody tell me how to modify an ITP file to ignore certain 
>> atoms (as
> if
>> they were not there) during simulations? I've changed charge and 
>> masses to
> 0.0
>> but I've got a lot of warnings during grompp preprocessing telling me 
>> that
>> those atoms has zero mass and the program stops after certain numbers 
>> of
>> warnings.
>>> Well you've still got bonded and LJ interactions acting to create 
>>> forces
>>> on particles with zero mass so it is no great surprise that things 
>>> are
>>> breaking. Probably what you want to do is go back to square one and 
>>> make
>>> topologies that don't have these atoms, but you haven't described 
>>> your
>>> chemistry, so we can't help you there.
> Mark
> Well, my idea is to "delete" atoms and then "undelete" them without 
> touching
> coordinates file. I guess that it could be done just touching topology 
> files.
> Is it correct??? The results I'm looking for is to turn invisible some 
> atoms.
> Any suggestion???
> Thanks in advance,
> César.-

More information about the gromacs.org_gmx-users mailing list