Hi Filip,
> Thus, if I want to simulate my systems with my previous parameters,and
> probably the most common, my limitation is 90K atoms (For the standard
> versions of GTX580 and GTX590). This means that I have to use only 8A
> water buffer. However, with ntt=1 will be much more. I will check this
> tomorrow.
To add some comment to this debate. The reason the benchmarks all focus on NTT=1 is because this is what most people (at least what the rough consensus was amongst the developers) should be using this for production dynamics. The typical protocol I use is to first minimize my system, then heat it to 150K or so using NVT with NTT=3 and then switch to NPT to allow the density to equilibrate where I heat it to 310K still using NTT=3. I then let this run for long enough that I believe the system has equilibrated away from the initial structure and then for the production at 310K I either run NVT or NPT but with ntt=1 and tautp=10.0. The long coupling constant of 10.0 for the Berendsen thermostat makes it similar to NVE and so adds the least perturbation to the system. As opposed to Langevin which scrambles all of your correlation functions etc.
Note, Berendsen has bad press because of all the coupling issues it has had. Hot solute, cold solvent etc. However, most of these problems are from the early days before PME. Additionally once your system has equilibrated the thermostat makes only minor corrections and so the issues with Berendsen should be negligible for a well equilibrated production run.
Hence most things in AMBER are mostly tuned for NVT (and NPT) with ntt=1.
All the best
Ross
/\
\/
|\oss Walker
---------------------------------------------------------
| Assistant Research Professor |
| San Diego Supercomputer Center |
| Adjunct Assistant Professor |
| Dept. of Chemistry and Biochemistry |
| University of California San Diego |
| NVIDIA Fellow |
|
http://www.rosswalker.co.uk |
http://www.wmd-lab.org/ |
| Tel: +1 858 822 0854 | EMail:- ross.rosswalker.co.uk |
---------------------------------------------------------
Note: Electronic Mail is not secure, has no guarantee of delivery, may not be read every day, and should not be used for urgent or sensitive issues.
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Sat Aug 20 2011 - 19:30:02 PDT
