[gmx-users] setting constraints to incomplete protein structures

Justin A. Lemkul jalemkul at vt.edu
Tue Oct 7 04:07:52 CEST 2008

wk yeo wrote:
> Hello,
> My responses are as follows:
>>> My questions:
>>> 1) Is it a good idea to use such catalytic domain structures for MD simulation? Or should I only use complete protein structures for MD?
>> That depends entirely upon what you are interested in simulating.
> I am interested in predicting the binding energy (via further processing using a docking software) of a ligand to a specific protein target based on the conformation of the binding pocket after MD.  How will missing residues in the other parts of the protein target affect the MD run for such purposes?

In biology, structure leads to function.  Missing residues may lead to spurious 
behavior, such as destabilization of the protein structure.

>>> 2) If I want to use Gromacs to conduct a MD of 0.1ps at 300K, do I need to constraint any atoms at the 'edges' during the MD run?
>> What kind of "edges?" Those that are adjacent to the missing segments? If
>> that's what you mean, see distance *restraints* or position *restraints*.
> Yes, I meant atoms that are adjacent to the missing segments. When should I use distance restraints and when should I use position restraints?

Distance restraints are what you are probably looking for.  You can fix the 
distance between the backbone atoms to the distance found in the crystal 
structure.  As for whether or not you can be confident in doing so is up to you.

Position restraints are typically applied during solvent equilibration to allow 
the solvent to orient around the solute.

> I have a new question: What if the protein-ligand complex structure obtained from PDB contains chloride ions? How should I handle any ions during the MD runs? Restraint them as well? Or should I just let them go through MD without any restraints?

If the ions were part of the precipitant or some sort of stabilizing co-solvent, 
they may be unnatural and therefore safe to ignore.  Adding (at least) 
counterions to an MD simulation is routine, so you may end up adding those ions 
back within during solvation.


> regards,
> wk yeo
> _________________________________________________________________
> Manage multiple email accounts with Windows Live Mail effortlessly.
> http://www.get.live.com/wl/all


Justin A. Lemkul
Graduate Research Assistant
Department of Biochemistry
Virginia Tech
Blacksburg, VA
jalemkul[at]vt.edu | (540) 231-9080


More information about the gromacs.org_gmx-users mailing list