On Wed, Sep 30, 2009 at 10:02 AM, Noel O'Boyle <baoille...@gmail.com> wrote: > 2009/9/29 Craig A. James <cja...@emolecules.com>: >> Tim Vandermeersch wrote: >>> >>> On Tue, Sep 29, 2009 at 10:50 PM, Noel O'Boyle <baoille...@gmail.com> >>> wrote: >>>> >>>> Nice work. >>>> http://timvdm.blogspot.com/2009/09/automorphism-group-bliss.html >>>> >>>> Does this mean you've solved the problem? >>> >>> Yes, but to continue towards correct canonical codes, we still need to >>> do some matrix math. The easiest solution would be to use Eigen 2 but >>> I'm not sure trunk should already depend on it... I can use git(hub) >>> for now. >>> >>> I think we should also remove the "break chiral ties" step. The new >>> FindStereogenicUnits function finds these stereocenters now regardless >>> of duplicated symmetry classes. This would mean symmetry classes are >>> topologically only (not regarding stereochemistry). >> >> This is correct. The "break chiral ties" step is only there because the >> initial symmetry analysis doesn't account for the chirality in the first >> place. Reading your series of blogs, it's clear that incorporating >> chirality into the symmetry analysis itself is a better algorithm. >> >> One question: it looks to me like the symmetry classes, and consequently the >> canonical labels, will be completely different using this new method, which >> means the canonical SMILES will all change. That's not a bad thing when >> weighed against the improved algorithm, just something to consider. > > I'm definitely in favour of the improved algorithm, but some other > things to consider from the bliss web site > (http://www.tcs.hut.fi/Software/bliss/index.html): "the computed > canonical form depends on the version and applied options of bliss". > Well, that's just a fact of life, I guess. Also "In principle, the > canonical forms should not depend on the computer platform in which > bliss is compiled and run, but this has not been extensively tested in > practice." This is a bit vague - how could the results be > architecture-dependent? Do we need to check for little-endian vs > big-endian bit string manipulations?
AFAIK this doesn't apply to the automorphism group. We don't actually use bliss to get the canonical labels, just to get the automorphism group. From there, we still have to assign canonical labels taking stereochemistry into account. Tim ------------------------------------------------------------------------------ Come build with us! The BlackBerry® Developer Conference in SF, CA is the only developer event you need to attend this year. Jumpstart your developing skills, take BlackBerry mobile applications to market and stay ahead of the curve. Join us from November 9-12, 2009. Register now! http://p.sf.net/sfu/devconf _______________________________________________ OpenBabel-Devel mailing list OpenBabel-Devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/openbabel-devel
