[gmx-users] my protein gets cut while preparing for em

[Tsjerk Wassenaar]

Hi Guillem,

First regarding the force field. You should not use the ffgmx one. That is
basically the old GROMOS87 force field, which has lots of errors in it.
Rather take one of the newer GROMOS96 force fields, or maybe an all atom
one, such as OPLS. That does pose you the problem of getting the right
parameters for the ligand.., sorry.

As for the rectangular rhombic dodecahedron.., that's explained in the
manual in Chapter 3 (periodic boundary conditions). Every box can be
represented as a triclinic box or a rectangular box with exactly equal
content. This also seems the cause of your molecule being split. Every part
is oriented such that the first atom is placed inside the rectangular box
you see. That has all to do with efficiency in the MD code. It does seem to
suggest though, that you have more subunits in the protein. Should it really
be one piece? How many *_?.itp files did you obtain after running pdb2gmx?
If there's more than one, while it is actually one chain, run pdb2gmx with
the option -merge

If some parts of this don't make sense, feel free to ask further... (but
also consider the manual and the user list archive).

Cheers,

Tsjerk


On 6/19/06, Olli Lehtonen <olehtone at focus.hut.fi> wrote:
>
>
>
> On Mon, 19 Jun 2006, Guillem Plasencia wrote:
>
> > Then i was curious and did trjconv my system (before running mdrun
> energy
> > minimization) into a pdb, and had a look at it, and i saw my protein was
> > splitted in three different molecules, one was inside the water box (but
> far
> > away from box center, lying in touch of one of the boxes' sides) and the
> > other two parts were completely outside the box.  Worst of all, they
> were no
> > more one single peptidic chain, but separated one.
>
> I was having somewhat similar problem recently when I had multiple
> peptides in my system (I used genconf to generate them) but in the final
> structure they all had the same chain identifier, so pdb2gmx thought that
> I have only one long peptide in the system (with some 1 nm long bonds).
>
> If this is the reason, one way to check how many chains there are is to
> interactively choose termini groups with pdb2gmx, you should be prompted
> for each peptide unit.
>
> By the way, if it is somehow possible to tell genconf to generate repeated
> units with different chain identifiers, I would be interested to know.
>
> regards,
>
> Olli
>
> _______________________________________________
> gmx-users mailing list    gmx-users at gromacs.org
> http://www.gromacs.org/mailman/listinfo/gmx-users
> Please don't post (un)subscribe requests to the list. Use the
> www interface or send it to gmx-users-request at gromacs.org.
> Can't post? Read http://www.gromacs.org/mailing_lists/users.php
>



-- 

Tsjerk A. Wassenaar, M.Sc.
Groningen Biomolecular Sciences and Biotechnology Institute (GBB)
Dept. of Biophysical Chemistry
University of Groningen
Nijenborgh 4
9747AG Groningen, The Netherlands
+31 50 363 4336
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://maillist.sys.kth.se/pipermail/gromacs.org_gmx-users/attachments/20060620/14b3f2e0/attachment.html>