Re: CpHMD: Warning when building atom of unknown mass (psfgen)

From: Brian Radak (brian.radak_at_gmail.com)
Date: Tue Jun 30 2020 - 18:18:30 CDT

Hi Aalap,

This is very probably an innocuous warning unless you want to use mass
repartitioning. The mass is "unknown" in the sense that it is not
present in the original PSF and therefore the code can only infer the
value that you intended (i.e. from the RTF). Upon restart the dummy
protons are present in the PSF and so the code can only assume that the
mass you provided is the mass that you want.

If you do wish to use mass repartitioning, then you must construct a PSF
with the dummy protons (e.g. by running a short CpHMD simulation) and
the performing repartitioning on that PSF and using it as the initial input.

HTH,

BKR

On 6/30/20 4:35 PM, Aalap Parikh wrote:
>
> Hello,
>
> I had a question about a particular warning message I have encountered
> during my usage of the CpHMD implementation in NAMD.
>
> The warning message appears when the initial dummy atom patches are
> applied at the beginning of the CpHMD run. These patches are defined
> in a CpHMD specific topology file (top_cph36_prot.rtf) which is
> distinct from the topology and parameters of the starting structure
> and are used to define the various possible protonation states of
> titratable residues. The dummy hydrogen mass is defined in the CpHMD
> specific parameter file (par_cph36_prot.prm) with the line:
>
> “MASS   131 HD     1.00800 ! polar H dummy”
>
> I am wondering why the WARNING! Message (shown in the psfgen excerpt
> below from the log file) is given. Specifically, why is the program
> seemingly unable to recognize the mass of the dummy atom, which is
> already defined in the parameter file? In what scenarios other than
> mass repartitioning might this warning lead to errors? How is the
> unknown mass ultimately assigned? I did not find much about this
> particular warning after a web search.
>
> TCL: namdcph) initializing constant pH PSF...
>
> psfgen) clearing structure, preserving topology and aliases
>
> psfgen) reading structure from psf file ../topology/ace_asp_nme.psf
>
> psfgen) reading coordinates, insertion codes, and element symbols from
> pdb file ../topology/ace_asp_nme.pdb
>
> psfgen) reading coordinates from namdbin file ../topology/ace_asp_nme.coor
>
> psfgen) reading velocities from velnamdbin file
> ../topology/ace_asp_nme.vel
>
> psfgen) applying patch ASPD to 1 residue(s)
>
> PROT:1
>
> TCL: WARNING! Building new atom PROT:1:HD1 with unknown mass.
>
> TCL:          This may cause issues, for example, with mass
> repartitioning.
>
> TCL: WARNING! Building new atom PROT:1:HD2 with unknown mass.
>
> TCL:          This may cause issues, for example, with mass
> repartitioning.
>
> psfgen) Info: guessing coordinates for 2 atoms (0 non-hydrogen)
>
> TCL: cphpsfgen) creating new alchemical side chain for PROT:1:ASP
>
> psfgen) applying patch ASPH1 to 1 residue(s)
>
> PROT:1
>
> TCL: cphpsfgen) merging alchemical sidechains for PROT:1:ASP
>
> psfgen) regenerating all angles
>
> psfgen) regenerating all dihedrals
>
> Also, when I restart the CpHMD simulation using the generated PSF
> (which now contains the dummy atoms), no warning is given:
>
> TCL: namdcph) initializing constant pH PSF...
>
> psfgen) clearing structure, preserving topology and aliases
>
> psfgen) reading structure from psf file 4.1/ace_asp_nme_prod0.psf
>
> psfgen) reading coordinates, insertion codes, and element symbols from
> pdb file 4.1/ace_asp_nme_prod0.pdb
>
> psfgen) reading coordinates from namdbin file 4.1/ace_asp_nme_prod0.coor
>
> psfgen) reading velocities from velnamdbin file 4.1/ace_asp_nme_prod0.vel
>
> psfgen) duplicate topology file ../../toppar/top_cph36_prot.rtf
>
> psfgen) applying patch ASPD to 1 residue(s)
>
> PROT:1
>
> psfgen) Info: guessing coordinates for 0 atoms (0 non-hydrogen)
>
> TCL: cphpsfgen) creating new alchemical side chain for PROT:1:ASP
>
> psfgen) applying patch ASPH1 to 1 residue(s)
>
> PROT:1
>
> TCL: cphpsfgen) merging alchemical sidechains for PROT:1:ASP
>
> psfgen) regenerating all angles
>
> psfgen) regenerating all dihedrals
>
> Thanks for the assistance,
>
> Aalap
>

This archive was generated by hypermail 2.1.6 : Thu Dec 31 2020 - 23:17:13 CST