[gmx-users] Re: pdb2gmx error

[Justin A. Lemkul]

On 6/27/12 2:42 PM, Shima Arasteh wrote:
> You are right.
>
> OK, the reason of adding FOR is what I explained a few minutes ago.
> All over, I'm supposed to use a modern FF . ( In similar researches I found CHARMM is suitable, so I would add FOR as the procedure explained in GROMACS.ORG)
>

This sounds like a good idea.

> For now, because I wanna regenerate an old simulation, I keep the FOR, then here it's stopped! Seeking for a solution!
>

There is one possible solution that I just thought of.  Rather than using FOR as 
its own residue, create a formylated valine, thus eliminating the problem with 
the call to CA of the preceding residue.

The .rtp entry:

[ FVAL ]
  [ atoms ]
     CN   CH1   0.380     0
     ON     O  -0.380     0
      N     N  -0.280     1
      H     H   0.280     1
     CA   CH1   0.000     2
     CB   CH1   0.000     3
    CG1   CH3   0.000     4
    CG2   CH3   0.000     5
      C     C   0.380     6
      O     O  -0.380     6
  [ bonds ]
     CN    ON
     CN     N
      N     H
      N    CA
     CA     C
      C     O
     CA    CB
     CB   CG1
     CB   CG2
  [ impropers ]
      N    CN    CA     H
     CN   -CA     N    ON
     CA     N     C    CB
     CB   CG2   CG1    CA

The .hdb entry:

FVAL    1
1   1   H   N   CN  CA

Then add FVAL to residuetypes.dat as a protein residue.  Processing the .pdb 
file you posted before (merging FOR and VAL into FVAL and adjusting names 
according to the .rtp entry above) works correctly.

Disclaimer: this method "works" (i.e. produces a topology) but I do not 
necessarily endorse the use of the parameters shown.

-Justin

-- 
========================================

Justin A. Lemkul, Ph.D.
Research Scientist
Department of Biochemistry
Virginia Tech
Blacksburg, VA
jalemkul[at]vt.edu | (540) 231-9080
http://www.bevanlab.biochem.vt.edu/Pages/Personal/justin

========================================