VMD-L Mailing List

From: Cindy Cregier (cindy.cregier_at_REMOVE_yahoo.com)
Date: Mon Mar 11 2024 - 20:03:24 CDT

 Hi Yixin,
I'm fairly new to NAMD, but your problem sounds similar to an issue I recently had.  I see that you're using parameters created from different time frames.  You'll find that the older parameter file has different atom definitions than the newer parameter file.  Atoms can be given one name in one parameter file and another name in the other parameter file.  Sometimes even numerical values given in the parameter files change between one generation and the next!  Also, look at the date of generation.  I found one file that had the same name, but a different generation date.  The contents had changed, but they kept the name of the topology file the same!  
Here's what I did.  I made sure that I used only the latest version of each topology and parameter file that's out there. I also made sure that the topology files used to create my psf and pdb files had the same definitions as the parameter files that I used for analysis.  Knowing what version of a topology file was used in the creation of your pdb and psf files helps a lot.  

I recommend opening the parameter files to see what's in them. Then, open your pdb and psf files that you've made.  The definitions in your pdb and psf files should help you determine what topology files were used to create them and therefore what parameter files you need to be using so that all of your atoms are defined.  Good luck!  
    On Thursday, March 7, 2024 at 03:27:19 PM EST, Josh Vermaas <vermaasj_at_msu.edu> wrote:
 
   Hi Yixin,
 
 Keep the list in the loop, so that the universe can learn about how to do this. If you have a psf/pdb pair for the membrane, and a psf/pdb pair for the graphene film, and you are happy with both of them, I usually recommend another topotools feature, the mergemols command.https://www.ks.uiuc.edu/Research/vmd/plugins/topotools/#TOC-Utility-Functions From what I recall, merge structures calls psfgen in the background, so in my hands it did some dumber stuff when it doesn't recognize something in the topology file. Topotools mergemols does not, and is basically like a fancy copy-paste command.
 
 -Josh
 
 On 3/7/24 1:43 PM, Yixin Chen wrote:
  
 Hi Josh,
 
 Thank you so much for your reply!  Following your guidance, I conducted tests on the direct simulation of the lipid membrane using the pdb and psf files generated by the VMD Membrane Builder tool directly, and it works. The problem should be in the step of converting membrane and graphene to psf. In the previous test, I used the following command in the Tcl console to convert PDB to PSF for both graphene film and lipid (POPC) membrane and combine the two using the tool(Merge structure) in VMD.
 
package require topotools
 topo retypebonds
 topo guessangles
 topo guessdihedrals mol reanalyze 0
 animate write psf name.psf
 
  Today, to get the correct psf, I attempted to use the provided command(in NAMD Tutorial, as shown) for generating psf files. While the generation was successful for the lipid membrane, I noticed that some information was lost in the process, resulting in an unusual appearance when viewed in VMD. For the graphene film, the conversion process encountered an error. Specifically, the error message indicated: "Error! Fail to recognize XXX (including CA, CD, N, NH... which are not included in the pdb) and unknown residue type GRA." I think that this error may be attributed to an incorrect topology file.
 package require psfgen topology top_all27_prot_lipid_na.inp segment U {pdb myfile.pdb} coordpdb  myfile.pdb U guesscoord writepdb  myfile .pdb writepsf  myfile .psf
 
 Could you please let me know how to correctly convert the pdb to psf for lipid membrane and graphene film and what topology file should be used. Thank you very much!!
 Best regards,
 Yixin
 
 
  On Wed, 6 Mar 2024 at 15:36, Vermaas, Josh <vermaasj_at_msu.edu> wrote:
  
    
Hi Yixin,
 
 
 
How did you build your psf? The lipid atomtypes should be something like CTL2 or CTL3 or something if they are in the lipid tail, not C12 or C13. Somehow your atomtype field in the psf is populated with atomnames instead, which don’t match atomtypes in the parameter file that you loaded. If you started from the membrane builder tool in VMD, it is possible that you loaded the pdb first into VMD, which sets atomtypes by default based on the atomname. Then when you did the merger, it is possible that the correct atomtypes were not transferred.
 
 
 
-Josh
 
 
  
From: <owner-vmd-l_at_ks.uiuc.edu> on behalf of Yixin Chen <yixin.chen1_at_ucdconnect.ie>
 Date: Wednesday, March 6, 2024 at 10:08 AM
 To: "vmd-l_at_ks.uiuc.edu" <vmd-l_at_ks.uiuc.edu>
 Subject: vmd-l: looking for help to solve FATAL ERROR： Didn't find vdw parameter for atom type C11
   
 
   
Dear all,
  
 
   
I hope this email finds you well. I am looking for help regarding an error that occurred when I ran my simulation in NAMD. FATAL ERROR： Didn't find vdw parameter for atom type C11.
   
 
   
My simulation model comprises a graphene film and a lipid (POPC) membrane, both of which were constructed using the Carbon Nanostructure and Membrane Builder tools in VMD. I have combined these two components into a single PDB and PSF file. However, upon running the simulation, the aforementioned error arises. Upon reviewing the PDB file, I noticed that C11 (and similarly C12, C13, etc.) is part of the lipid membrane structure. 
   
 
   
Initially, I thought that I might miss a parameter file, so I included all the necessary topology files in the configuration file. Despite this, the error persists. the parameter file I used is shown:
  
  
parameters          par_all27_prot_lipid_na.inp
   
parameters          par_all22_prot.prm
   
parameters          par_all36_carb.prm
   
parameters          par_all36_cgenff.prm
   
parameters          par_all36_lipid.prm
   
parameters          par_all36_na.prm
   
parameters          par_all36_prot.prm
   
parameters          par_amber2charmm.inp
  
  
 
   
Additionally, I attempted to modify the atom designation from C11 to C in the PDB, PSF, and reference files, but the error message persists with a change in atom type (e.g., "Didn't find vdw parameter for atom type C12").
   

 If anyone has encountered a similar issue or has expertise in troubleshooting NAMD simulations, I would greatly appreciate any insights or guidance you can provide. Thank you very much for your attention and help!
   
 
   
Best regards,
   
Yixin
      
     
 
 --
Josh Vermaas

vermaasj_at_msu.edu
Assistant Professor, Plant Research Laboratory and Biochemistry and Molecular Biology
Michigan State University
vermaaslab.github.io