[gmx-users] emtol

[K.A. Feenstra]

Nguyen Hoang Phuong wrote:
> 
> 
> On Fri, 17 Jan 2003, [iso-8859-1] Tanos C. C. Fran?a wrote:
> 
> 
>>	Hi folks,
>>
>>	What is considered a suitable tolerance (emtol) for a energy minimization
>>of a big dimeric protein  (884 residues) in a water box  ?  In the article
>>cited bellow, Berendsen said that he performed an em with steepest descent
>>until a tolerance of 0,01 Kj/mol.nm. How to do this ? In my machine the most
>>I achieved was 1000 Kj/mol.nm in a water box. The calculation always stops
>>when the Fmax is about 1000 and gives me the following mesage: "Step size
>>too small. Converged to machine precision". What does it means ?
>>
> 
> 
> I usually perform em for systems of protein and water and I can reach to
> Fmax = 0.001 using cg with emstep = 10^{-6} and double precision. Usually
> the annealing simulations are done and then the em are performed for the
> obtained configurations.

It also depends on what you want to do. For equilibration before starting
md, an Fmax of around 1000 is perfectly fine. For normal modes, and perhaps
other (analysis) methods, CG is necessary.

The initial stepsize is not too important; mdrun will adjust it up or
down depending on the results it got on the previous steps. If you use
constraints (e.g. Shake, Lincs, or Settles in SPC), you can forget good
minimization; you *can* still minimize well enough to start a run (at
least, it usually works fine), but no way for normal modes and the like.


-- 
Groetjes,

Anton
  ________ ___________________________________________________________
|        | Anton Feenstra                                            |
| .      | Dept. of Pharmacochemistry - Vrije Universiteit Amsterdam |
| |----  | De Boelelaan 1083 - 1081 HV Amsterdam - The Netherlands   |
| |----  | Tel: +31 20 44 47608 - Fax: +31 20 44 47610               |
| ' __   | Feenstra at chem.vu.nl - http://www.chem.vu.nl/afdelingen/FAR|
|  /  \  |-----------------------------------------------------------|
| (    ) | Dept. of Biophysical Chemistry - University of Groningen  |
|  \__/  | Nijenborgh 4 - 9747 AG Groningen - The Netherlands        |
|   __   | Tel +31 50 363 4327 - Fax +31 50 363 4800                 |
|  /  \  | K.A.Feenstra at chem.rug.nl - http://md.chem.rug.nl/~anton   |
| (    ) |-----------------------------------------------------------|
|  \__/  | "Why do *you* use Constraints?" (H. J. C. Berendsen)      |
|________|___________________________________________________________|