[gmx-users] Re: About Gromacs

[Tsjerk Wassenaar]

Hi Xi Ouyang,

> >   Right now I'm doing wavelet-based coarse-graining model of stretching
> the protein. Wavelet is easy to play with under Matlab. Now I want to do is
> try to do wavelet decomposition (like fourier decomposition) on the
> potential function of the protein and try to do the MD, using the decomposed
> potential. ( In Gromacs language, is potential contained in the force
> field?? )

Am I right to interpret your problem as attempting to rewrite the
complex potential function in a limited set of wavelets? I think this
comes close to normal modes analysis (see manual), where you want to
replace the complex potential with a small number of the normal modes.
I can see a difficulty in connecting the dynamics of the particles to
these 'wavelets'. But also mind that proteins have considerable
anharmonic behaviour, which might not connect to well with wavelets.

> And comparing the computing time cost between the original model
> and the decomposed model. But now I cannot find the interface of the wavelet
> into gromacs.
> >     Because I am a EE guy, I know nothing about MD and gromacs. Installing
> the gromacs under linux took me 2 weeks! So I really appreciate your help if
> you can give me some advice. That will really save me a lot of time.
> >

Two weeks? That might be a record ;)

>
>  First do some MDS background reading - such as the first few chapters of
>  the GROMACS manual. MDS are intrinsically atomistic, so you will need to
>  either cast your wavelet decomposition in such a fashion to use GROMACS,
>  or write custom software.
>
>  Mark
>

Especially read about forces, force fields and their connection to
motions of paritcles in MD.

Cheers,

Tsjerk

-- 
Tsjerk A. Wassenaar, Ph.D.
Junior UD (post-doc)
Biomolecular NMR, Bijvoet Center
Utrecht University
Padualaan 8
3584 CH Utrecht
The Netherlands
P: +31-30-2539931
F: +31-30-2537623