From: Peter Freddolino (petefred_at_ks.uiuc.edu)
Date: Wed Mar 23 2011 - 20:19:09 CDT
For an RBCG system you need to set the terminal patches to NONE, or else
add NTER/CTER patches to your topology file (we didn't use charged
termini in the papers where rbcg was applied, but they should be easy;
follow the logic of the atomistic nter/cter with a Qd bead for the nter
backbone and a Qa bead for the cter backbone).
However, if namd now supports martini, there's really no earthly reason
to use this model unless you want to specifically compare with previous
results using the same force field.
On 03/23/2011 02:37 PM, Michael Zimmermann wrote:
> Peter and NAMD Users,
> The steps I took in trying the RBCG builder were as follows:
> 1) download 1UBQ.pdb
> 2) Manually remove solvent atoms
> 3) make psf with VMD (v.1.9): Extensions -> Modeling -> Automatic PSF
> - Default topology file used - top_all27_prot_lipid_na.inp
> 4) Use this psf+pdb for RBCG
> - in VMD, Extensions -> Modeling -> CG Builder
> - used the cg beads as defined in the cgtools plugin directory
> 5) Now, I need a psf for the cg system
> - In VMD, Extensions -> Modeling -> Automatic PSF Builder
> - Remove default topology file. Add the topology file from the
> cgtools plugin directory
> - Run autopsf
> The last step in 5) has errors about the terminal patches used - they
> are not defined. The ubiquitin files that I have used are in the
> attached zip file. My goal is to use the RBCG procedure on larger
> systems, but I have been using ubiquitin as a testing structure.
> Any help would be appreciated.
> On Tue, Mar 22, 2011 at 9:17 PM, Peter Freddolino <petefred_at_ks.uiuc.edu
> <mailto:petefred_at_ks.uiuc.edu>> wrote:
> Hi Michael,
> There should no longer be any need for a special version of autopsf; the
> links you posted are to an old stopgap version that fixed a bug (which
> has now been corrected in the main versions).
> Could you elaborate on how you tried to build the RBCG models? In order
> to help, we need a concrete series of steps from 1ubq.pdb to your error,
> because when I try doing this it works fine.
> Regarding MARTINI, I know a couple of grad students in Klaus's group
> were working on it, but I'm afraid I don't know the current status...
> perhaps a current group member could comment.
> On 03/22/2011 04:11 PM, Michael Zimmermann wrote:
> > NAMD Users,
> > I want to do RBCG simulations on some large systems, but have been
> > unable to build RBCG models for small proteins (like 1UBQ) due to
> > "/Error: failed on end of //segment; molecule destroyed./" errors.
> > These are errors for applying the NTER and CTER patches. If I try to
> > set these to NONE then autopsf also fails. Following the advice of:
> > http://www.ks.uiuc.edu/Research/namd/mailing_list/namd-l/11194.html
> > and
> > http://www.ks.uiuc.edu/Research/vmd/mailing_list/vmd-l/14725.html
> > I have tried to source a different version of autopsf, but I don't see
> > how to do so. I have tried to at the tk console, but get:
> > versions provided for package "autopsf": 1.3, then 1.2/"
> > If anyone has further advice on making RBCG models, or a working RBCG
> > model that I could learn from, I would appreciate it.
> > Also, is there any news about using MARTINI in NAMD
> > (http://www.ks.uiuc.edu/Research/namd/mailing_list/namd-l/8802.html)?
> > I am using VMD v.1.9 Windows OpenGL, CUDA on 64-bit WinXP.
> > --
> > Michael Zimmermann
> > Ph.D. student in Bioinformatics and Computational Biology
> > Department of Biochemistry, Biophysics and Molecular Biology
> > Iowa State University
> Michael Zimmermann
> Ph.D. student in Bioinformatics and Computational Biology
> Department of Biochemistry, Biophysics and Molecular Biology
> Iowa State University
This archive was generated by hypermail 2.1.6 : Wed Feb 29 2012 - 15:56:48 CST