Re: Re: Re: [AMBER] how do MM_PBSA without water

From: Jason Swails <>
Date: Wed, 21 Apr 2010 07:55:36 -0400


2010/4/21 geyan <>

> Hi,Bill,sorry to reply you a little late.
> My system is a complex composed of several moleculars of protein and one
> molecular of DNA.Because the past study has clearly demonstrated that they
> are interacting with each other without water due to the spore enviroment,
> so I run the simulation in vacuum,without water,in other words,No solvent.

You did not even use implicit solvent (i.e. GB)? Note that while explicit
water molecules may not be involved in enzyme activity, the solvent almost
surely plays a role in modulating the secondary and tertiary structure of
your system, so to neglect it altogether has to be done very carefully. At
the very least, Amber charges are derived using a protocol that accounts for
increased charge polarization that is typical of polar solvents, so the
default charge model is not necessarily appropriate for a gas-phase
simulation. Running gas phase simulations properly are more involved than
simply changing flags in the sander/pmemd MDIN file, so take care with this.

> Also,I know the MMPBSA calculations used solvent effect,as you
> say,explicit solvent. But to my special system,I have to turn it

MM/PBSA does not use explicit solvent. It uses an implicit solvent model to
account for solvation effects. You could conceivably change igb from
whatever value it's set to to 0 in the _MMPBSA_gb.mdin input file generated
by MMPBSA (see the -make-mdins and -use-mdins flags in the manual).
However, I'm pretty sure this will fail with a python "IndexError" when it
tries to parse non-existent solvation terms (ESURF/EGB).

As a side note, MM/PB(GB)SA stands for Molecular Mechanics, Poisson
Boltzmann (or Generalized Born) Surface Area. Vacuum simulations make no
use of PB, GB, or solvent accessible surface area (which is used to model
the non-polar solvation energy), so this method does not seem readily
appropriate to your situation (if indeed vacuum simulations are what you're
really interested in).

> off,I don't know whether it will affect the calculation or not?what I'm now
> more interested is how to change the prameters to turn it off,so that it can
> produce a more reasonable result?. Regarding the,I have no idea
> with it.I will have a try on google.
> Best regards.Thank you.
> You will not find much about it on google, as it is quite new and still in
early stages of development.

Good luck!

Jason M. Swails
Quantum Theory Project,
University of Florida
Ph.D. Graduate Student
AMBER mailing list
Received on Wed Apr 21 2010 - 05:00:04 PDT
Custom Search