[gmx-users] Minimization and torsion angles

[Mark Abraham]

On 8/02/2011 2:47 AM, abdullah ahmed wrote:
> Hello,
>
> I have an input structure with poor torsion angles and after 
> minimization with gromacs (both steepest descent and congugate 
> gradient) I found that they were not corrected. There is enough free 
> space for the torsion angles to be corrected.
> (I am sure of this because when I run the same structure in Insight it 
> is able to correct them. Both programs give very similar structures 
> with the major difference being the torsion angles.)
>
> Has this happened because Gromacs potentials do not penalize poor 
> omega angles?

Peptide bonds ordinarily have improper dihedrals on them to keep them 
close to planar.

> Or have I done something wrong? (I have added by .mdp file at the end 
> of the mail)
> Is there a way besides applying a restraint to every torsion angle to 
> induce Gromacs to correct them?

EM is a local minimization method, and the result depends on the method 
and its parameters. Perhaps choosing a larger step size with 
steepest-descent algorithm will fix your issues.

Mark

>
> Thank you in advance for you help,
> Abdullah
>
>
>
> ;    User spoel (236)
> ;    Wed Nov  3 17:12:44 1993
> ;    Input file
> ;
> cpp                 =  /usr/bin/cpp
> define              =  -DPOSRES
> constraints         =  none
> integrator          =  cg
> nsteps              =  2000
> ;
> ;    Energy minimizing stuff
> ;
> emtol               =  2
> emstep              =  0.01
>
> nstcomm             =  1
> ns_type             =  grid
> rlist               =  1
> rcoulomb            =  1
> rvdw                =  1
> Tcoupl              =  no
> Pcoupl              =  no
> gen_vel             =  no
>

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://maillist.sys.kth.se/pipermail/gromacs.org_gmx-users/attachments/20110208/45c4cd5b/attachment.html>