[gmx-users] Long range Lennard Jones

[Justin Lemkul]

On 8/29/13 1:18 AM, Gianluca Interlandi wrote:
> Justin,
>
> I respect your opinion on this. However, in the paper indicated below by BR
> Brooks they used a cutoff of 10 A on LJ when testing IPS in CHARMM:
>
> Title: Pressure-based long-range correction for Lennard-Jones interactions in
> molecular dynamics simulations: Application to alkanes and interfaces
> Author(s): Lague, P; Pastor, RW; Brooks, BR
> Source: JOURNAL OF PHYSICAL CHEMISTRY B Volume: 108 Issue: 1 Pages: 363-368
> Published: JAN 8 2004
>

I cannot say for sure whether the DispCorr implementation in Gromacs is the same 
or comparable to IPS in CHARMM.  You would have to test that, and also test a 
more complex system than simple liquids.

> There is also a paper by Piana and Shaw where different cutoffs for non-bonded
> are tested with CHARMM22 on Anton:
>
> http://www.plosone.org/article/info%3Adoi%2F10.1371%2Fjournal.pone.0039918
>
> They found some subtle differences, in particular for cutoffs shorter than 9 A.
> However, Anton uses abrupt truncation (no switching) and I believe that the
> differences they found at cutoffs > 9 A would be much smaller if they had used a
> finer mesh (as they show at the 8 A cutoff). I always use fourierspacing=1.0
>

Note that Shaw's group is using their own modified force field, CHARMM22*, so 
that factors in here, as well.  I do recall that paper, though I haven't read it 
in a while, so details may be fuzzy.  Wasn't the principal point to test methods 
for long-range electrostatics and the influence of cutoffs in that context?  I 
seem to recall LJ taking a backseat there.  It is certainly true that 
short-range Coulomb cutoffs are more flexible when using Ewald methods.

> I agree though that it strongly depends on the system and I have always run
> control simulations but never found significant differences in the case of just
> proteins.
>

That's good to know that you've tested different setups.  Are the water 
interaction energies comparable?  Have you tested model compounds or just full 
proteins?  Over how long of a time period?

> Finally, I have not tested it in gromacs, but in NAMD there is a performance
> gain of 25% when using the shorter cutoff. This is a factor to consider. When I
> asked for Teragrid supercomputing allocations back in 2006 and 2007 and I
> suggested 10/12/14 cutoff, the reviewers always complained and cut my requested
> time of 20% with the justification that I must use a shorter cutoff.
>

That's unfortunate.  In my opinion, it is shameful that desired performance 
exceeds proven integrity of the data.  If all we're after is performance, I can 
write up a dozen papers in the next few months with completely useless data, but 
they will be produced quickly! ;)

-Justin

-- 
==================================================

Justin A. Lemkul, Ph.D.
Postdoctoral Fellow

Department of Pharmaceutical Sciences
School of Pharmacy
Health Sciences Facility II, Room 601
University of Maryland, Baltimore
20 Penn St.
Baltimore, MD 21201

jalemkul at outerbanks.umaryland.edu | (410) 706-7441

==================================================