[MMTK] Problems with Normal Mode analysis

[Pawel Kedzierski]

  Dear All,

  First of all, I wish You everything good in the New Year!

  And then, I would like to ask for some help ;-)

  I am trying to run Normal Mode analysis using MMTK on several
PDB files, using Python 2.0.1, Numeric 22, Scientific 2.4.1,
netcdf 3.5 and MMTK 2.2 on Linux 2.4.18. The calculations are
run in a loop and if there are problematic cases they should be just
counted and skipped but they should not crash nor hang the process.
I experience two problems:

  1. Large memory consumption. For some files its 700-900MB and
     the calculations slow down to a crawl as the machine has 512MB RAM;
   Q: Is there a way to limit/estimate memory requirements? Ideally I'd like
      to force MMTK to do the same calculation using less memory
      (compromising time), but a clean way to give up if the requirements
      are too high would also be welcome. I tried setting timeouts using
      signal module but the MMTK code seem to ignore signals?  Can I, for
      example, estimate the memory requirements from e.g. nbasis parameter?

  2. Random segfaults which cannot be caught using Python exception
     handling.
   Q: Can I prevent them? Or, at least, does anybody know how to debug
     them (noninteractively at best, since they usually happen after
     several hours of calculations...)

The code was borrowed from the deformation_modes.py example:

universe         = InfiniteUniverse(DeformationForceField())
universe.protein = Protein(plik.name, model='calpha')
nbasis           = max(10, universe.numberOfAtoms()/5)
cutoff, nbasis   = estimateCutoff(universe, nbasis)
if cutoff is None:
  modes = NormalModes(universe)
else:
  subspace = FourierBasis(universe, cutoff)
  modes = SubspaceNormalModes(universe, subspace)

Greetings,

     Pawel Kedzierski

-- 
--> email: P.Kedzierski@mml.ch.pwr.wroc.pl
    Molecular Modelling Laboratory
    Institute of Physical and Theoretical Chemistry I-30
    Wroclaw University of Technology
    Wroclaw, POLAND