[gmx-users] Longer scale simulations

[Anton Feenstra]

Tsjerk Wassenaar wrote:

> 
> Hi,
> 
> Depending on the geometry of your system and the flexibility, it can be 
> worth performing the simulation in a molecular shaped box. This may save 
> you 50% on computational time. If you like you can contact me off the list.
> 
> Ran Friedman wrote:
> 
>> Well, that's not very encouraging, but it's better than being wrong!

In addition, you can use dummified hydrogen atoms. This removes the highest
frequency motions allowing the timestep to be raised from 2fs to 4-5fs
giving an additional performance gain of a factor of 2 without noticable
loss of accuracy. I've described this in J. Comput. Chem. 20 (8),
pp786-798, 1999, and David and Erik have rigorously tested it recently
(but I don't have that reference at hand though, sorry). It is also
already built into pdb2gmx (and grompp), simply by adding '-dummy hydrogens'
when you run pdb2gmx.


-- 
Groetjes,

Anton
  _____________ _______________________________________________________
|             |                                                       |
|  _   _  ___,| K. Anton Feenstra                                     |
| / \ / \'| | | Dept. of Pharmacochem. - Vrije Universiteit Amsterdam |
|(   |   )| | | De Boelelaan 1083 - 1081 HV Amsterdam - Netherlands   |
| \_/ \_/ | | | Tel: +31 20 44 47608 - Fax: +31 20 44 47610           |
|             | Feenstra at chem.vu.nl - www.chem.vu.nl/~feenstra/       |
|             | "If You See Me Getting High, Knock Me Down"           |
|             | (Red Hot Chili Peppers)                               |
|_____________|_______________________________________________________|