Re: [AMBER] On the significant dG error in the MMGBSA results.dat

From: Aron Broom <broomsday.gmail.com>
Date: Tue, 2 Dec 2014 18:35:47 -0500

you should be able to see from the MMPBSA output log, where the errors are
coming from. I believe the NMA code will actually also warn you about
likely problems (assuming you are using the lastest AmberTools). For
instance, typically you want in the range of 1 snapshot per degree of
freedom. So, depending on whether you had hydrogens fixed or not, you
could be looking at ~2-3 times the number of non-water atoms, which I
imagine for your system is going to mean you want in the range of 10,000
snapshots. But, you also want those snapshots to be independent of each
other, that is, not correlated. So you probably can't go lower than taking
a snapshot every 10ps or so. Which, oddly enough, would mean about 100ns,
which is what you have. So, yes, you could improve your error by recording
snapshots as frequently as say 10ps. This will also, by default, reduce
the error from the enthalpy portion of the calculation, since the error
will decrease with the square-root of the number of measurements (assuming
again that they are independent, and also that you are in fact exploring
within the global energy minimum).

On Tue, Dec 2, 2014 at 4:29 PM, James Starlight <jmsstarlight.gmail.com>
wrote:

> Dear Amber users!
>
> Performing several mmgbsa calculations (with NMA for enthropy calculations)
> for my test systems I've noticed significant error corrections in the dG
> found in final result. From the below log you can find that the order of
> such corrections in both cases are equal to the calculated values. Here I
> used 100 ns trajectory of the production run of the membrane protein system
> as the input for mmgbsa; totally using 50 snapshots both for the entalpy
> (each snapshot has been taken every 250000 steps) as well as for the
> entropy calculations.
> 1) Does the enthropy term make main contribution to the errors?
> 2) Is it possible to reduce its value (e.g using longest MD
> trajectories,choosing smaller sampling intervals or do better minimization
> as well as method for hessian diagonalization in case of NMA)?
>
>
>
> -------------------------------------------------------------------------------
> For system1
>
> -------------------------------------------------------------------------------
>
> -------------------------------------------------------------------------------
> Using Normal Mode Entropy Approximation: DELTA G binding = -8.1087 +/-
> 5.8570
>
> -------------------------------------------------------------------------------
>
> -------------------------------------------------------------------------------
> For system 2
>
> -------------------------------------------------------------------------------
>
> -------------------------------------------------------------------------------
> Using Normal Mode Entropy Approximation: DELTA G binding = -10.3566 +/-
> 7.7316
>
> -------------------------------------------------------------------------------
>
> -------------------------------------------------------------------------------
>
>
> Thanks for suggestions,
>
> James
> _______________________________________________
> AMBER mailing list
> AMBER.ambermd.org
> http://lists.ambermd.org/mailman/listinfo/amber
>



-- 
Aron Broom M.Sc
PhD Student
Department of Chemistry
University of Waterloo
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Tue Dec 02 2014 - 16:00:02 PST
Custom Search