[gmx-users] unusually long run required

[Anton Feenstra]

Jaimin Shah wrote:
> Dear Dr. Spoel,
> 
> As per you instruction, I set the time step of 0.002 ps in the pr.mdp file
> and nstxout = 250000. That necessitated setting the no. of steps to
> (nstep) to 1250000. Upon mdrunning, I get the message that the run will be
> completed in July 2006. Am I right in doing the settings?

That seems excessively long. 2.5 ns usually takes a few days at most. 
What type of system are you simulating, and which hardware do you use?

> However, I thought that if I want a conformation extracted at every 500
> ps, then nstep should be equal to 5. Setting it that way does not give me
> any .trr file.

You want a frame every 500 ps, and have a 0.002 ps per timestep, so:
500 ps / (0.002ps/step) = 250000 steps

Be aware that in this way, for 2.5 ns simulation, you will only get 
(2.5ns/500ps)+1 = 6 conformations. Is that what you want?

-- 
Groetjes,

Anton

* NOTE: New Phone & Fax numbers (below) *

  _____________ _______________________________________________________
|             |                                                       |
|  _   _  ___,| K. Anton Feenstra                                     |
| / \ / \'| | | Dept. of Pharmacochem. - Vrije Universiteit Amsterdam |
|(   |   )| | | De Boelelaan 1083 - 1081 HV Amsterdam - Netherlands   |
| \_/ \_/ | | | Tel: +31 20 59 87608 - Fax: +31 20 59 87610           |
|             | Feenstra at few.vu.nl - www.few.vu.nl/~feenstra/         |
|             | "If You See Me Getting High, Knock Me Down" (RHCP)    |
|_____________|_______________________________________________________|