Hi Noel, As you know, I'm working on a distance geometry builder for Open Babel, like in RDKit. The bonus would be stereo-correct coordinate generation and likely faster than the current OBBuilder.
I sat down and wrote a unit test, but there's something totally weird. Here's the PR: https://github.com/openbabel/openbabel/pull/1738 So the test tries to do something like this: obabel -:'CN([C@@H]1C(=O)C(=C([C@@]2([C@H]1C[C@@H]1C(=C(O)c3c([C@@]1(C)O)c(Cl)ccc3O)C2=O)O)O)C(=O)N)C' -ocan --gen3d dg The test appears to pass. But then I generate a mol2 file and round-trip back to canonical SMILES: CN([C@@H]1C(=O)C(=C([C@@]2([C@H]1C[C@H]1C(=C(O)c3c([C@]1(C)O)c(Cl)ccc3O)C2=O)O)O)C(=O)N)C Whoops - the 4th tetrahedral carbon should be C@@H but is now C@H and the last one should be C@@ but is now C@. (I've attached the mol2 file - looks normal in Avogadro) So I'm looking at the stereo: > Old: OBTetrahedralStereo(center = 7, viewFrom = 6, refs = 25 8 27 , clockwise) > New: OBTetrahedralStereo(center = 7, viewFrom = 6, refs = 8 27 25 , clockwise) > Old: OBTetrahedralStereo(center = 8, viewFrom = 7, refs = H 9 2 , clockwise) > New: OBTetrahedralStereo(center = 8, viewFrom = 7, refs = 2 H 9 , clockwise) > Old: OBTetrahedralStereo(center = 10, viewFrom = 9, refs = H 16 11 , > clockwise) > New: OBTetrahedralStereo(center = 10, viewFrom = 9, refs = 11 H 16 , > clockwise) Wait, those don't seem to line up at all, but the stereo configs are considered to match by OB? Can you help me wrap my brain around the comparisons of tetrahedral configs? Thanks, -Geoff
test10.mol2
Description: chemical/mol2
------------------------------------------------------------------------------ Check out the vibrant tech community on one of the world's most engaging tech sites, Slashdot.org! http://sdm.link/slashdot
_______________________________________________ OpenBabel-Devel mailing list OpenBabel-Devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/openbabel-devel
