[gmx-users] missing residues

[Nuno R. L. Ferreira]

Hi Yu

On Wednesday 17 March 2004 05:33 pm, Yu Sun (Denis) wrote:
> Dear gmx-users,
>
> I am dealing with a large protein (~ 75kD). There are some missing
> residues and atoms. For example, there are 9 residues missing in sequence (
> from
> 21 to 29 ). I used InsightII to add all the missing residues and atoms.

Perhaps you could try this.
After adding the missing residues, try to perform an energy minimization only 
on the added residues.  Then a little md run, say 20 ps, freezing only the 
protein without the 9 residues ( or with a high value of position restraints 
force), letting this residues to accomodate in their surroundings. After that 
minimize all the system, and you are ready for a run.

If these new residues are in the surface, I think that this is quite easy to 
do. If its in the core could be more difficult to perform. Perhaps some md 
with an increase in the vdw radius during the run, till it reaches the final 
radius values for the atoms. 

> Then I am trying to use Gromacs to do energy minimization. But I got  fmax
> = inf  and energy is very high( positive 10^5).
>
> I took out residues from 20 to 30  to obtain a pdb file with 11
> residues and fixed  residue 20 and 30 to do energy minimization (in
> vacuum), still got very high positive energy. 

Because you don't let the protein to leave from the starting state, which 
could have bad Vdw contacts.

When not fixing 20 and 30,
> negative energy can be obtained ( negative 10^3).
>

Nuno