VMD-L Mailing List

From: Tristan Croll (tristan.croll_at_qut.edu.au)
Date: Thu Nov 17 2011 - 01:02:52 CST

Joakim,

Definitely not impossible - you just have to make a patch. As an example, I just bodged together the following patch for the glutamic acid-to-N-terminus case (it's easier to patch starting from a glutamine residue, but the end result is identical):

PRES LIG3 0.00000 ! linkage for cyclic peptide
                ! 1 refers to the side chain of a glutamine residue
                ! 2 refers to the N terminus
                ! use in a patch statement, perform initial
                ! generation using first NONE
DELETE ATOM 1NE2
DELETE ATOM 1HE21
DELETE ATOM 1HE22
BOND 1CD 2N
ANGLE 1CD 2N 2CA 1CA 1C 2N
ANGLE 1OE1 1CD 2N 1CD 2N 2HN
DIHE 1CD 2N 2CA 2C 1CD 2N 2CA 2HA 1CD 2N 2CA 2CB
DIHE 1HG1 1CG 1CD 2N 1CB 1CG 1CD 2N 1HG2 1CG 1CD 2N
DIHE 1CG 1CD 2N 2HN 1CG 1CD 2N 2CA
DIHE 1OE1 1CD 2N 2HN 1OE1 1CD 2N 2CA
DIHE 1CG 1CD 2N 2CA
IMPR 2N 1CD 2CA 2HN 1CD 1CG 2N 1OE1

Add that somewhere in your top_all27_prot_lipid_na_test.inp file. When going through AutoPSF, before the "create chains" step hit "Edit Chain" and set the patch on the N terminus to "NONE". Then create the chains, and hit "Add Patch". Under "Patch Type" put "LIG3". The glutamine residue is residue 1, and the N-terminal residue is residue 2. I've checked, and this will successfully create the correct structure. To model it, you'll have to add a few missing bond, angle and dihedral parameters to the par_all27_prot_lipid_na.inp file. Ideally you should optimise these, but for that you should get pretty close just by copying the equivalent parameters corresponding to backbone amides.

My approach to working out which parameters are missing is generally to just try to start a NAMD simulation, and read the resulting error message. Just keep adding the missing parameters until it runs.

Hope this helps.

Cheers,

Tristan

From: owner-vmd-l_at_ks.uiuc.edu [mailto:owner-vmd-l_at_ks.uiuc.edu] On Behalf Of Joaim Swedberg
Sent: Thursday, 17 November 2011 9:42 AM
To: vmd-l_at_ks.uiuc.edu
Subject: vmd-l: Patching Glu/N-terminus peptide bond type bond

Hi all,
I have spent the last two weeks trying to manually produce topology and Psf files that will allow for cyclisation between Glu acid side chain and the N-terminus as well as through Lys and the C-terminus without getting anything to work. I begin to doubt if it is possible with the current nomenclature of +N and -C to refer to bonds to the previous and next residue. Has anybody else tried this with success?

Regards

Joakim Swedberg

Research Officer
Prof David Craik's Group
The Institute for Molecular Bioscience
The University of Queensland
Brisbane QLD 4072
Australia

Phone: +61 (0)413 165 805
Email: j.swedberg_at_imb.uq.edu.au
Web: http://www.imb.uq.edu.au/