AMBER Archive (2009)

Subject: Re: [AMBER] Thermodynamic Integration calculation on an artificial anion receptor

Date: Thu Oct 29 2009 - 05:35:24 CDT


> I am trying to calculate the relative binding free energies of an
> artificial receptor with various anions in explicit solvent by means
> of TI. Firstly, I remove the partial charges on the anion, followed by
> a mutation to another anion and finally putting the charges on the
> mutated anion. However, when removing the charges on the anion, the
> anion no longer binds to the receptor. Do I have to put constraints on
> the complex or are there any other ways to curb this problem? Any
> comment or suggestion is greatly appreciated.

Adding constraints is a way to solve this, but make sure to account for
them correctly. Check e.g. in

Wang, J.; Deng, Y.; Roux, B. Absolute Binding Free Energy Calculations
Using Molecular Dynamics Simulations with Restraining Potentials. Biophys.
J., 2006, 91, 2798-2814.

You are making life hard for yourself by removing a negative charge in one
step and then adding it back in another. Maybe the neutral species really
doesnt bind to the receptor and thats what you see. Is there a way you can
keep the intermediate compound also negatively charged? That may help
convergence a lot. Also, Amber11 (due in Spring) will include a way to
conduct such transformations in a single step, so the ligand would stay
negative throughout the transformation.

Kind Regards,


Dr. Thomas Steinbrecher
BioMaps Institute
Rutgers University
610 Taylor Rd.
Piscataway, NJ 08854

AMBER mailing list