Re: AMBER: Amber 9: Generating Input files for EVB simulations of the crosslinked structures (Hydrogels)

From: David A. Case <case.scripps.edu>
Date: Thu, 8 Nov 2007 12:16:57 -0800

On Thu, Nov 08, 2007, sanket deshmukh wrote:

> Thanks for your suggestions David. I created a pdb file without any
> united atom in it (attaching it with this email) but still I am facing
> the same problem. Antechamber can't find the connectivity properly.

Two points:

1. Antechamber on works on single residues, not on entire chains or sets of
chains.

2. You pdb file has lines like these:

ATOM 1 C3 GLY 1 6.000 12.000 10.000
ATOM 2 C3 GLY 1 5.000 13.000 11.000
ATOM 3 C GLY 1 5.000 11.000 9.000
ATOM 4 O GLY 1 6.000 12.000 8.000
ATOM 5 OS GLY 1 4.000 12.000 8.000
ATOM 6 C3 GLY 1 3.000 11.000 9.000
ATOM 7 C3 GLY 1 5.000 13.000 9.000
ATOM 8 OS GLY 1 4.000 14.000 10.000

First, the coordinates look like garbage. Second, atom names within a residue
must have unique names, and you have several atoms in GLY 1 named "C3".
Third, since what you are calling GLY clearly isn't glycine, Amber will most
likely get confused: don't call something GLY that is not glycine.

I could go on (e.g. you have another GLY residue (starting at atom 212) that
contains only hydrogens and nothing else). You need to "start small", and
make sure you understand how antehcamber works with a minimal molecule, then
think about building up more complex systems.

And, use a molecular viewer (say, VMD or Chimera) to visualize your pdb file
before sending it to antechamber: this will certainly help avoiding obvious
flaws in the pdb file.

...dac

-----------------------------------------------------------------------
The AMBER Mail Reflector
To post, send mail to amber.scripps.edu
To unsubscribe, send "unsubscribe amber" to majordomo.scripps.edu
Received on Sun Nov 11 2007 - 06:07:32 PST
Custom Search