[gmx-users] all-atom pca

gromacs query gromacsquery at gmail.com
Tue May 12 00:44:37 CEST 2015


Hi Tsjerk

In -proj plot of v1 vs Time I used extreme values of V1 and saw structures
on those times. When I compare these structures to the structures obtained
from -extr the atomic positions (of 'same' extremes) are not 'exactly'
matching even after fitting. I tried with double precision as well.

Am not sure why they are not exactly fitting (trjconv_d -fit). Its like if
I try to fit same structures they should exactly overlap. May the fitting
algorithm has a glitch when molecule is big as in my case (~ 500 atoms)


Thanks
Jio

On Mon, May 11, 2015 at 10:08 PM, Tsjerk Wassenaar <tsjerkw at gmail.com>
wrote:

> Hi Jio,
>
> You'll have to look at the projections (-proj), and see at which time these
> have an extreme value.
>
> Cheers,
>
> Tsjerk
>
> On Mon, May 11, 2015 at 11:00 PM, gromacs query <gromacsquery at gmail.com>
> wrote:
>
> > Hi Tsjerk
> >
> > Thanks for reply.
> >
> > >> So you can use the times
> > I am unable to use it as explained below.
> >
> > I get extremes output pdb like this:
> >
> > g_covar -s avg.pdb -f all_frames.pdb
> > g_anaeig -s avg.pdb -extr two_extremes.pdb -first 1 -last 1 -nframes 2
> >
> >
> > Visually two structures in two_extremes.pdb looks fine and expected as
> well
> > but the times mentioned in two_extremes.pdb are MODEL 0 t =0 and   MODEL
> 1
> > t =1 which seems not correct. In my case t = * should be a frame number
> as
> > I am not using .tpr file and frames are fed as pdb file in which MODEL 0
> to
> > n is each frame.
> >
> > Thanks
> >
> > On Mon, May 11, 2015 at 9:47 PM, Tsjerk Wassenaar <tsjerkw at gmail.com>
> > wrote:
> >
> > > Hi Jio,
> > >
> > > The scores for all-atom structures will be strongly correlated to those
> > for
> > > heavy-atom structures. So you can use the times for the heavy-atom
> > extreme
> > > projections to extract all-atom structures.
> > >
> > > Cheers,
> > >
> > > Tsjerk
> > >
> > > On Mon, May 11, 2015 at 10:32 PM, gromacs query <
> gromacsquery at gmail.com>
> > > wrote:
> > >
> > > > Hi All
> > > >
> > > > I am doing PCA and its working fine if I choose few atoms only
> (heavy)
> > > and
> > > > I get extreme pdb file having heavy atoms for, say, eigenvector 1.
> But
> > I
> > > > want to have extremes of all-atom structures, so I thought of doing
> PCA
> > > on
> > > > all-atoms but I end up with having Segmentation default error which I
> > > think
> > > > related to comp memory issue.
> > > > I think (please correct me) the extreme structures (heavy atoms)
> would
> > > > belong somewhere in the trajectory file. If I can get frame numbers
> of
> > > > these extremes calculated on heavy atoms then I can use same frame
> > > numbers
> > > > to dump all-atom structures. But how to get frame numbers for these
> > > > extremes? may be RMSD with respect to extremes on heavy atoms.
> > > >
> > > > Thanks
> > > > Jio
> > > > --
> > > > Gromacs Users mailing list
> > > >
> > > > * Please search the archive at
> > > > http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before
> > > > posting!
> > > >
> > > > * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists
> > > >
> > > > * For (un)subscribe requests visit
> > > > https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users
> or
> > > > send a mail to gmx-users-request at gromacs.org.
> > > >
> > >
> > >
> > >
> > > --
> > > Tsjerk A. Wassenaar, Ph.D.
> > > --
> > > Gromacs Users mailing list
> > >
> > > * Please search the archive at
> > > http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before
> > > posting!
> > >
> > > * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists
> > >
> > > * For (un)subscribe requests visit
> > > https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or
> > > send a mail to gmx-users-request at gromacs.org.
> > >
> > --
> > Gromacs Users mailing list
> >
> > * Please search the archive at
> > http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before
> > posting!
> >
> > * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists
> >
> > * For (un)subscribe requests visit
> > https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or
> > send a mail to gmx-users-request at gromacs.org.
> >
>
>
>
> --
> Tsjerk A. Wassenaar, Ph.D.
> --
> Gromacs Users mailing list
>
> * Please search the archive at
> http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before
> posting!
>
> * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists
>
> * For (un)subscribe requests visit
> https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or
> send a mail to gmx-users-request at gromacs.org.
>


More information about the gromacs.org_gmx-users mailing list