[gmx-users] NMA

[Anton Feenstra]

Bert de Groot wrote:
> Ruben Martinez Buey wrote:
> 
>>Hi all,
>>Thanks alot for your reply,
>>I only want to see global motions of the bacbone or C-alpha , starting from a
>>2.7 A. resolution structure, so maybe it is impossible to minimize until  Fmax < 0.0001?
> 
> 
> no it should still be possible, but the nearest minimum may
> be far away (especially if you run in vacuo). You can check how
> the structure changes during the EM. In the ideal case the structure
> shouldn't drift too much. If it does, then that's probably a sign
> of a poor starting structure, or a forcefield/vacuum artefact.
> 
> 
>>and maybe it can work with shorter minimizations?
> 
> 
> perhaps. as I wrote before, it may be worth a try.
> IIRC, I got fastest convergence for the EM when I started with steep and
> then switched to cg, or even when doing iterative cycles of steep and cg.

There's even an option for that in the .mdp: nstcgsteep, to do one
'steep' step every so many cg steps.


-- 
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)                               |
|_____________|_______________________________________________________|