From: Josh Vermaas (vermaas2_at_illinois.edu)
Date: Tue Apr 19 2016 - 16:27:38 CDT

Hi Olya,

Thats a weird error, and one I haven't seen before. What solvate is
supposed to do is read in your original psf pdb pair, construct a
sufficiently large waterbox, put the two together, and carve out waters
that are too close to the solute. Unfortunately, it does so by calling
psfgen, which apparently doesn't like something about how the system is
put together when it reads in the psf and thinks you don't have a
residue 1 of segment A, and refuses to set coordinates accordingly. What
you could do to avoid this is to build a waterbox alone, merge the two
systems together using topotools, and remove the waters using an
atomselection:

set sel [atomselect top "(not water) or (water and not same residue as
(name OH2 and within 2.4 of (not water)))"]
$sel writepsf blah.psf
$sel writepdb blah.pdb

The trouble with this approach is that you'll face the same problem
during ionization (probably), so you should probably figure out why
psfgen is misbehaving. Does something simple like:

readpsf ferritin.psf
coordpdb ferritin.pdb

writepsf test.psf
writepdb test.pdb

work at all? Or does this also throw errors?

-Josh Vermaas

On 04/19/2016 02:59 PM, Olya Kravchenko wrote:
> Hi all,
>
> I am trying to solvate ferritin (it is a spherical molecule that
> consists of 24 identical chains, 66456 atoms ) in a water box and keep
> getting the same error regardless of whether I run it in a tcl/tk
> console or graphic interface. The error is only for the first 9
> residues of the first chain, total of 151 atoms. When I run solvate,
> i.e.
>
> package require solvate
> solvate ferritin.psf ferritin.pdb -t 5 -o solvated
>
> I see warnings such as these:
>
> psfgen) no residue 1 of segment A
> psfgen) Warning: failed to set coordinate for atom N THR:1 A
> psfgen) no residue 1 of segment A
> psfgen) Warning: failed to set coordinate for atom HN THR:1 A
> psfgen) no residue 1 of segment A
> psfgen) Warning: failed to set coordinate for atom CA THR:1 A
> psfgen) no residue 1 of segment A
> psfgen) Warning: failed to set coordinate for atom HA THR:1 A
> psfgen) no residue 1 of segment A
> psfgen) Warning: failed to set coordinate for atom CB THR:1 A
> ..
> ..
>
> and so on for the next 151 atoms in a ferritin.pdb file.
>
> When the file solvated.pdb is created, the first 151 atoms have zero
> coordinates:
>
> ATOM 1 N THR A 1 0.000 0.000 0.000 -1.00 0.00 A
> ATOM 2 HN THR A 1 0.000 0.000 0.000 -1.00 0.00 A
> ATOM 3 CA THR A 1 0.000 0.000 0.000 -1.00 0.00 A
> ATOM 4 HA THR A 1 0.000 0.000 0.000 -1.00 0.00 A
> ATOM 5 CB THR A 1 0.000 0.000 0.000 -1.00 0.00 A
> ..
> ..
>
> In the ferritin.pdb file the same atoms have coordinates and and look
> normal in VMD before I solvate the structure:
>
>
> ATOM 1 N THR A 1 29.972 78.277 67.795 1.00 0.00 A
> ATOM 2 HN THR A 1 30.255 79.195 67.517 0.00 0.00 A
> ATOM 3 CA THR A 1 30.644 77.930 69.037 1.00 0.00 A
> ATOM 4 HA THR A 1 30.049 78.354 69.834 0.00 0.00 A
> ATOM 5 CB THR A 1 30.648 76.421 69.061 1.00 0.00 A
> ..
> ..
>
> I don't understand what causes this error and why only on the first 9
> residues of the very first chain?
>
> I searched the archives, but did not find a similar issue being
> discussed. Initially I thought there may be non-standard atom or
> residue names that do not correspond to those in the parameter files,
> but all seems to be consistent, and in any case I do not see errors of
> this sort for the rest of the molecules even though the chains are
> identical (and so are atom and residue names). I use
> top_all36_prot.rtf for topology.
>
> I'd very much appreciate your input!
>
> Olga
>