Hi Dan,
Wow, thanks for being so diligent in tracking this bug down! I have been
slowly whittling down my trajectories due to finite memory issues and was
just about to begin post-processing the atommaps again, so your fix is
perfectly timed. Thanks!
Take care,
Samuel Bowerman
Ph.D. Candidate
Department of Physics
Illinois Institute of Technology
On Tue, Sep 20, 2016 at 9:06 AM, Daniel Roe <daniel.r.roe.gmail.com> wrote:
> Hi,
>
> I was able to track down the bug - in 'crdaction' where the action
> modified the topology, the wrong frame was being passed in leading to
> memory issues. I've fixed the issue in GitHub and will work on a patch
> for AmberTools. If you use the GitHub version the script previously
> posted should work fine now. Thanks for the report and the files!
>
> -Dan
>
> On Wed, Aug 31, 2016 at 9:12 AM, Daniel Roe <daniel.r.roe.gmail.com>
> wrote:
> > Hm - sounds like a potential bug. Would you mind sending me off-list some
> > files I could use to reproduce this behavior (sounds like trajectories
> only
> > need to be 2 frames to see the problem).
> >
> > I am currently in the middle of a move so I may not be able to get to
> this
> > as soon as I would like, but if there is a problem I'd eventually like to
> > correct it.
> >
> > Glad to hear you could still do the analysis you wanted.
> >
> > -Dan
> >
> >
> > On Tuesday, August 30, 2016, Samuel Bowerman <
> sbowerma+amber-l.hawk.iit.edu>
> > wrote:
> >>
> >> Hi Dan,
> >>
> >> Thanks for your response. While I agree that what you've posted
> *should*
> >> work as is, it unfortunately does not. The first frame of the
> trajectory
> >> is reordered appropriately, but subsequent frames are riddled with
> strange
> >> atom configurations: for example, the atoms within C-J get lumped into a
> >> giant ball instead of being mapped to their original locations. I am
> not
> >> too familiar at working with trajectories as COORDS sets, so I can't
> offer
> >> a guess as to why this may be occurring.
> >>
> >> That being said, my most pressing need is to use cpptraj to identify
> >> different structural clusters in groups A and B, so I've decided to just
> >> strip my data to these groups and then use atommap on the resulting
> >> trajectories prior to clustering. While being able to use a wider
> variety
> >> of cpptraj's utilities on the full simulations would be fantastic, this
> >> will satisfy my original plans for these trajectories.
> >>
> >> Thanks again for your assistance.
> >>
> >> Take care,
> >> Samuel Bowerman
> >> Ph.D. Candidate
> >> Department of Physics
> >> Illinois Institute of Technology
> >>
> >> On Fri, Aug 26, 2016 at 3:40 PM, Daniel Roe <daniel.r.roe.gmail.com>
> >> wrote:
> >>
> >> > On Tue, Aug 23, 2016 at 3:23 PM, Samuel Bowerman
> >> > <sbowerma+amber-l.hawk.iit.edu> wrote:
> >> > >
> >> > > If I inspect the atommap.dat file, then I get the expected Target
> <->
> >> > > Reference index mapping for fragments 1 and 2, but fragments 3
> through
> >> > > 10
> >> > > are filled with "--- ---" for target indices and names. I believe
> >> > > this
> >> > > may be a result of those fragments not requiring mapping, but I have
> >> > > also
> >> > > found a two-year old thread
> >> > > (http://archive.ambermd.org/201504/0141.html
> >> > )
> >> > > that mentioned system size potentially being an issue - my system is
> >> > > roughly 30,000 atoms.
> >> >
> >> > Oh boy - yeah that's way bigger than any system I've tested 'atommap'
> >> > on. It was originally developed to re-map small drug-like molecules. I
> >> > don't remember offhand but I think the largest I ever tested on was
> >> > about 128 atoms or so. I could probably enhance the performance a lot
> >> > by doing things on a per-residue basis but there was no real demand
> >> > for it when I was originally writing the code and unfortunately I
> >> > haven't had a lot of free time.
> >> >
> >> > If the fragments themselves are contiguous and all that has changed is
> >> > the ordering, i.e. the original is:
> >> >
> >> > A B C ...
> >> >
> >> > and the new order is
> >> >
> >> > B A C ...
> >> >
> >> > what you might be able to do is manually reorder whichever trajectory
> >> > (or trajectories) are shortest via 'strip' and 'combinecrd'. For
> >> > example, say I wanted to reorder the first to match the second. I
> >> > would do something like this (I'm oing to use COORDS data sets so I'm
> >> > assuming you have enough memory):
> >> >
> >> > parm All.parm7
> >> > # Load entire trajectory as COORDS set A
> >> > loadcrd All.nc name A
> >> > # Strip everything but A from COORDS set A
> >> > crdaction A strip !<mask for A>
> >> > # Repeat for B
> >> > loadcrd All.nc name B
> >> > crdaction B strip !<mask for B>
> >> > # Repeat for C-J
> >> > loadcrd All.nc name CJ
> >> > crdaction CJ strip !<mask for C-J>
> >> > # Combine
> >> > combinecrd B A CJ crdname Reorder
> >> > # Write out
> >> > crdout Reorder Reorder.nc
> >> >
> >> > Haven't tested the above but I think it will work. Hope this helps,
> >> >
> >> > -Dan
> >> >
> >> > >
> >> > > If anyone could suggest an explanation, a fix, or a work-around for
> >> > > this
> >> > > problem, I would greatly appreciate it. Thanks!
> >> > >
> >> > > Take care,
> >> > > Samuel Bowerman
> >> > > Ph.D. Candidate
> >> > > Department of Physics
> >> > > Illinois Institute of Technology
> >> > > _______________________________________________
> >> > > AMBER mailing list
> >> > > AMBER.ambermd.org
> >> > > http://lists.ambermd.org/mailman/listinfo/amber
> >> >
> >> >
> >> >
> >> > --
> >> > -------------------------
> >> > Daniel R. Roe, PhD
> >> > Department of Medicinal Chemistry
> >> > University of Utah
> >> > 30 South 2000 East, Room 307
> >> > Salt Lake City, UT 84112-5820
> >> > http://home.chpc.utah.edu/~cheatham/
> >> > (801) 587-9652
> >> > (801) 585-6208 (Fax)
> >> >
> >> > _______________________________________________
> >> > AMBER mailing list
> >> > AMBER.ambermd.org
> >> > http://lists.ambermd.org/mailman/listinfo/amber
> >> >
> >> _______________________________________________
> >> AMBER mailing list
> >> AMBER.ambermd.org
> >> http://lists.ambermd.org/mailman/listinfo/amber
> >
> >
> >
> > --
> > -------------------------
> > Daniel R. Roe, PhD
> > Department of Medicinal Chemistry
> > University of Utah
> > 30 South 2000 East, Room 307
> > Salt Lake City, UT 84112-5820
> > http://home.chpc.utah.edu/~cheatham/
> > (801) 587-9652
> > (801) 585-6208 (Fax)
>
>
>
> --
> -------------------------
> Daniel R. Roe
> Laboratory of Computational Biology
> National Institutes of Health, NHLBI
> 5635 Fishers Ln, Rm T900
> Rockville MD, 20852
> https://www.lobos.nih.gov/lcb
>
> _______________________________________________
> AMBER mailing list
> AMBER.ambermd.org
> http://lists.ambermd.org/mailman/listinfo/amber
>
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Tue Sep 20 2016 - 08:30:02 PDT
