From: Pawel Kedzierski (pawel.kedzierski_at_pwr.edu.pl)
Date: Wed May 19 2021 - 15:00:34 CDT

W dniu 19.05.2021 o 19:04, Arthur Pereira da Fonseca pisze:
>
> Hello,
>
> I’m trying to run a dynamic analysis using NAMD. First I generate a
> valid psf file using VMD. For the two actions I need a valid parameter
> file to work with my molecule. The problem is: I'm running a peptide,
>  from pdb server,  that contains engineered residues. That means I
> have to provide specific parameter files in order to run these.
>
> So my question is: how can I generate these files?
>
> I tried with CGenFF, but I got an error saying: "unfulfilled valence
> in aromatic subgraph".
>
Indeed, your molecule has unfulfilled valences everywhere cause it does
not have hydrogen atoms. The automated parametrization services require
a chemically valid structure. Apart from that, the file you attached
contains water which is not part of the peptide and you should remove it
before parametrization.

A simple solution which may work for you is to download Avogadro
(https://urldefense.com/v3/__https://avogadro.cc__;!!DZ3fjg!o7VFZEs3XyfmgDMlc9t-V9XJ_N1kCV4OtlF6FTwYmwMmPEDRuKrNSJn0rBftI-lt-Q$ ) and do the following with your PDB file:

  * Select→"Select solvent", then press Del to remove water
  * Build→"Add hydrogens to pH" to add hydrogens automatically. Still
    the result should be carefully checked and possibly corrected.
  * File→Save as, select "Sybyl mol2" as the file type and use this for
    parametrization.

Note that this way the entire peptide will be treated as single residue
and it will not match the original PDB so you will need to replace it.

However, given that you try to parametrize an invalid structure, I have
doubts that your state of understanding is sufficient to do this
correctly. Parametrization is not a black box procedure and you should
not expect the software to "know chemistry"; you are responsible for
good science here. Therefore my best advice is to ask someone knowing
how to do it to guide you, this is likely a topic for many hours and not
for email communication.

>  I also tried reading their FAQ and using the
> softwares they suggest to generate the mol2 files, but with no luck. I
> tried using SPARTAN, from the "Parameterizing a Novel Residue"
> tutorial. But, again, I found myself with an error.
>
> So is there any other way to generate the needed files? Or something I
> can do to overcome these errors? I know it must be a problem with
> carbon valances within the peptide, but I don’t have enough chemical
> knowledge to fix this. Anyway, I'm attaching my molecule to this message.
>
Well, you need the chemical knowledge plus the knowledge on employed
modeling techniques to achieve meaningful results. Software will
faithfully calculate whatever you setup even if that doesn' make any sense.

Paweł

> Any help is appreciated.
>
> Thanks!
>
> Arthur
>