[gmx-users] Removing PBC from Replica Exchange Trajectory

[Robert Johnson]

Is there any way to remultiplex a REMD trajectory? If so, this chain
of events should work:

1. Demultiplex the REMD trajectory (with demux.pl) to obtain a
continuous trajectory
2. Use trjconv -pbc nojump on the demultiplexed trajectories
3. Remultiplexed the "nojumped" trajectories back

I tried carrying out step 3 using: trjcat -f
<demultiplex-nojump-trajectories> -o
<remultiplexed-nojump-trajectories> -demux replica_index.xvg

However, I don't think this worked. Even though the ssDNA remained
whole in these "remultiplexed-nojump-trajectories", the structures
didn't always match those from the original REMD trajectories. Any
idea on how this could be done?
Bob



On 3/15/07, Yang Ye <leafyoung81-group at yahoo.com> wrote:
> Thanks for Tsjerk's input. Rob, please make sure that that portion of
> DNA is always binding together. Since you have a ssDNA, just center the
> center residue.
> mol was a non-existing option. It sounds meaningful =) Just use whole,
> please.
>
> Regards,
> Yang Ye
>
> Tsjerk Wassenaar wrote:
> > Hi Robert,
> >
> > Maybe it's a bit nasty, but you may try the following, assuming that
> > your ends stay a bit together:
> >
> > 1. Translate your system such that the last atom of chain A is
> > centered in the _rectangular_ unitcell.
> > 2. Generate a .tpr file from the translated system
> > 3. Convert the .tpr to .pdb
> >
> > Gromacs places molecules in the box based on the first atom. Since
> > you're dealing with DNA, your chains will be head to tail. Placing the
> > tail of chain A in the center would mean that there's a periodic image
> > of your chain B which has its first atom in the center. This image
> > will be chosen during the generation of the .tpr file. Converting the
> > .tpr to .pdb should give both chains close to each other. Processing
> > the trajectory will be another thing...
> >
> > Tsjerk
> >
> >
> > On 3/15/07, Robert Johnson <bobjohnson1981 at gmail.com> wrote:
> >> I made an index file that contained the middle portion of the DNA
> >> strand. I then used the -center option to center this middle portion
> >> in the box and then outputted the entire molecule. However, the
> >> molecule still appears broken. What is this -pbc mol option? That
> >> doesn't seem to be one of the available pbc options.
> >> Bob
> >>
> >> On 3/14/07, Yang Ye <leafyoung81-group at yahoo.com> wrote:
> >> > A rather simple but effective way is to make an index which
> >> contains 1/4
> >> > of your protein file. Use trjconv with -center tric -pbc whole/mol to
> >> > center that 1/4 of your protein and output the whole protein.
> >> >
> >> >
> >> > Tsjerk Wassenaar wrote:
> >> > > Hi Bob,
> >> > >
> >> > > You could try -pbc cluster. If that doesn't work, it'll be
> >> difficult.
> >> > > The option -pbc nojump will only work on continuous trajectories.
> >> > >
> >> > > Cheers,
> >> > >
> >> > > Tsjerk
> >> > >
> >> > > On 3/14/07, Robert Johnson <bobjohnson1981 at gmail.com> wrote:
> >> > >> Hello everyone,
> >> > >> I'm trying to visualze the conformations of a ssDNA molecule
> >> obtained
> >> > >> from a
> >> > >> REMD trajectory. As expected, there are parts of the simulation
> >> where
> >> > >> the
> >> > >> oligomer is broken due to the PBC wrapping. I've been trying to fix
> >> > >> this with
> >> > >> trjconv. However, the -pbc nojump option doesn't work - it actually
> >> > >> makes the
> >> > >> broken portions of the molecule worse and gives a terribly
> >> distorted
> >> > >> trajectory. Since the REMD trajectory isn't continuous (i.e.
> >> there are
> >> > >> instaneous jumps in the coordinates associated with REMD swaps),
> >> the
> >> > >> nojump
> >> > >> option may have problems. Does anyone know of any other way or
> >> other
> >> > >> tools that
> >> > >> I could use to reassemble my broken DNA strand?
> >> > >> Thanks,
> >> > >> Bob Johnson
> >> > >> _______________________________________________
> >> > >> 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
> >> > >>
> >> > >
> >> > >
> >> >
> >> > _______________________________________________
> >> > 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
> >> >
> >> _______________________________________________
> >> 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
> >>
> >
> >
>
> _______________________________________________
> 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
>