[gmx-developers] Running energy minimization using ga

[David van der Spoel]

On Wed, 27 Mar 2002, Anton Feenstra wrote:

>If I understand you correctly, you generate a number of different
>structures using your genetic algorithm, and then you want to have
>energy values for these structures. This is a way you could do it:
>1) use grompp one time, to generate a topology that describes your
>molecule (assuming that your genetic algoritm does not change the
>molecule, only the conformation)
>2) put all your generated structures in a 'trajectory', this could
>simply be a series of .pdb files combined into one
>3) using the topology from step 1 and the 'trajectory' from step 2,
>use the '-rerun' option in mdrun to calculate energies for all 
>structures in your 'trajectory'.
>4) read the energies, either from the energy file, or from the log
>file that mdrun produces.

Actually it could work like this, since one always has a population of say
100 conformations. After energies have been computed for all of these
the genetic algorithm kicks in to do mixing etc. and the generates a new
traj (series of structures). That can be used with the same tpr file
ofcourse. The energy calculation would even be quite efficient.

Groeten, David.
________________________________________________________________________
Dr. David van der Spoel, 	Biomedical center, Dept. of Biochemistry
Husargatan 3, Box 576,  	75123 Uppsala, Sweden
phone:	46 18 471 4205		fax: 46 18 511 755
spoel at xray.bmc.uu.se	spoel at gromacs.org   http://zorn.bmc.uu.se/~spoel
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++