From: Peter Freddolino (petefred_at_umich.edu)
Date: Fri Oct 30 2020 - 08:00:10 CDT

Hi Mark,
A couple of thoughts here.
For autopsf -- autopsf is primarily intended to be used through the GUI,
because if you want the control of a command line application, you should
probably be using psfgen directly anyway. The autopsf GUI allows you to set
any terminal patches that you want (eg 5PHO/3PHO on the RNA chains), but
this functionality is not easily exposed in the command line interface.

Regarding the CYS SG coordinates, that is very odd; can you send a minimal
working example? This may be caused by the fact that autopsf is also trying
to automatically deprotonate cysteines when needed based on proximity to a
metal ion, which may cause double patching in your case.

For the topology files used by psfgen, I wonder if the charmm27 files are
left over in your environment for some other reason. Could you try
psfcontext new delete
right before you start loading the topologies?

Also, for any vmd questions it is very useful to specify the version number
that you are running.
Best,
Peter

On Thu, Oct 29, 2020 at 8:58 PM White, Mark <mawhite_at_utmb.edu> wrote:

> Hello,
>
> I found a strange difference in behaviour between the VMD packages PSFGEN
> and AUTOPSF.
>
> I can use AUTOPSF to generate a TIP3 solvated and SOD/CLA ionized PDB/PSF
> file. However, it does not apply the CYSD patch by deleting the CYS HG1
> atom, but by setting the atom positions to {0,0,0}.
>
> If I use PSFGEN then it does not use the CHARM26 atom types HB1 and HA2,
> but HB and HA. It appears to use the top_all27_prot_lipid_na.inp TOPPOLOGY
> file even though the CHARMM26 files are specified and recognized
>
> I am running the following script in VMD:
>
> package require psfgen
> package require readcharmmtop
> package require solvate
> package require autoionize
>
> lappend topfiles [file join $env(CHARMMTOPDIR) top_all36_prot.rtf]
> lappend topfiles [file join $env(CHARMMTOPDIR) top_all36_lipid.rtf]
> lappend topfiles [file join $env(CHARMMTOPDIR) top_all36_na.rtf]
> lappend topfiles [file join $env(CHARMMTOPDIR) top_all36_carb.rtf]
> lappend topfiles [file join $env(CHARMMTOPDIR) top_all36_cgenff.rtf]
> lappend topfiles [file join $env(CHARMMTOPDIR)
> toppar_all36_carb_glycopeptide.str]
> lappend topfiles [file join $env(CHARMMTOPDIR)
> toppar_water_ions_namd.str]
>
> pdbalias ...
>
> segment AP1 {pdb A04.pdb
> first NTER
> last CTER}
> patch CYSD AP1:5
> patch CYSD AP1:8
> patch CYSD AP1:26
> patch CYSD AP1:29
> patch CYSD AP1:50
> patch CYSD AP1:55
> patch CYSD AP1:72
> patch CYSD AP1:16
> patch CYSD AP1:19
>
> segment B01 {pdb B04.pdb
> auto none}
> segment DNA1 {pdb D04.pdb
> first 5PHO
> last 3PHO }
> segment ENA2 {pdb E04.pdb
> first 5PHO
> last 3PHO }
> segment Z LSB04a.psf01 {pdb Z04.pdb}
> segment M01 {pdb M04.pdb}
> regenerate angles dihedrals
> coordpdb A04.pdb AP1
> coordpdb B04.pdb B01
> coordpdb D04.pdb DNA1
> coordpdb E04.pdb ENA2
> coordpdb M04.pdb M01
> coordpdb Z04.pdb Z01
> guesscoord
> writepdb LSB04.pdb
> writepsf LSB04.psf
>
> As you can see, I have a ligand, two RNA chains, Mg, and Zn atoms the
> later which require the CYSD patch.
> The equivalent AUTOPSF script for VMD is:
>
> mol new LSB04.pdb
> autopsf -mol 0 -solvate -ionize -prefix LSB04a -patch CYSD AP1:5 -patch
> CYSD AP1:8 -patch CYSD AP1:16 -patch CYSD AP1:19 -patch CYSD AP1:26 -patch
> CYSD AP1:29 -patch CYSD AP1:50 -patch CYSD AP1:55 -patch CYSD AP1:72
> solvate LSB04a.psf LSB04a.pdb -t 25 -o LSB04aw25
> autoionize -psf LSB04aw25.psf -pdb LSB04aw25.pdb -sc 0.15 -o LSB04aw25i
> writepdb LSB04aw25i.pdb
> writepsf LSB04aw25i.psf
>
> What is the issue with PSFGEN? Is the toppology 'INP' hard encoded to be
> top_all27_prot_lipid_na.inp ?
> Also AUTOPSF will only use 3TER and 5TER MeO on the RNA not 3PHO, 5PHO
> terminations.
>
> Best regards,
> Mark
>
> Mark Andrew White, Ph.D.
> Associate Professor of Biochemistry & Molecular Biology,
> Manager, SCSB Macromolecular X-ray Laboratory
> 6.658A Basic Sciences, UTMB, Galveston, TX
> http://xray.utmb.edu
>