AMBER Archive (2007)

Subject: Re: AMBER: large RMS fluctuations in turning off VDW interactions by TI

From: David A. Case (case_at_scripps.edu)
Date: Mon Jul 02 2007 - 15:20:27 CDT

On Mon, Jul 02, 2007, David Mobley wrote:
>
> I agree with Julien here, and also recommend the paper he mentioned,
> as well as the Beutler and van Gunsteren paper (referenced therein) on
> softcore potentials from about 1994. You might also have a look at
> this preprint on a study on essentially this issue:
> http://www.dillgroup.ucsf.edu/~dmobley/papers/steinbrecher.pdf (the
> Beutler and van Gunsteren paper is also referenced there).
>
> The next version of AMBER will have "soft core potentials"
> implemented; these are designed to deal with this problem by removing
> the singularity in dV/dlambda and reducing the numerical noise.

I feel I should point out that Table 1 and Fig. 2 of the paper mentioned above
show that you should be able to get good results, with low statistical
uncertainties, using the klambda option of Amber8 or 9. The advantages of the
soft-core potentials (at least in this study) are visible, but relatively
minor. Thomas is travelling right now, but maybe when he reads this, he can
post the RMS fluctuations from this studies.

Which values of lambda are giving the high RMS flucutations (or is it at all
lambda values)? Also, the code is mostly set up to handle disappearing atoms
-- you might want to see if the high RMS is correlated with having appearing
atoms (or with anything else).

....dac

-----------------------------------------------------------------------
The AMBER Mail Reflector
To post, send mail to amber_at_scripps.edu
To unsubscribe, send "unsubscribe amber" to majordomo_at_scripps.edu