On Fri, Aug 13, 2010, Ben Roberts wrote:
> >> The first is that both sleap and tleap seem to create spurious bonds
> >> between the two nickel ions, which are far enough apart (3.5 to 4 Ã )
> >> that they shouldn't be bonded to each other.
The problem is in the way the metal ion library files are created. The
distance between the atoms is irrelevant.
> loadamberprep 1FWJ_chg2.prep
This creates a residue with "head" and "tail" atoms, so it will create bonds
to neighboring residues unless that is turned off with TER cards.
Note that the standard ions Amber uses have null heads and tails, which
explains why you were not getting bonds between Na+ ions, even though your
PDB file treats Na+ and NI1 in equivalent ways.
To see this, type the following commands in tleap
loadoff ions08.lib
desc Na+
loadamberprep 1FWJ_chg2.prep
desc NI1
You will see the difference.
So, Martin's scripts (or whatever you are using) either should create ions
in a way analgous to how Amber handles ions (see
amber11/dat/leap/lib/ions08.cmd), or arrange for TER cards to be put into the
pdb files.
>
> I was loath to send the PDB file straightaway, it being a 3 MB file.
Since only two lines of the pdb file were relevant, you could have
experimented to find a small example that still illustrated the problem.
(See, e.g., below).
Note to Bill (and others):
tleap *does* read and use CONECT records. Try playing around with the
following PDB file (e.g. with and without the CONECT record):
ATOM 1 Na+ Na+ 1 54.825 84.332 92.969
ATOM 2 Na+ Na+ 2 57.060 83.121 90.539
CONECT 1 2
...dac
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Sat Aug 14 2010 - 06:30:04 PDT
