Greg-
Both the chirality-AddHs() and memory leak fixes to RDKit seem to be
working perfectly by my tests. Great work and thanks again for being
so quick with responses to questions and updating RDKit with solutions!
-Marshall
On Apr 14, 2009, at 9:43 PM, Greg Landrum wrote:
Marshall,
On Tue, Apr 14, 2009 at 11:35 PM, Marshall Levesque
<marsh...@emolecules.com> wrote:
Here is the file and the output of a short test with the molecules,
which
match visual inspection of produced 3D structures via AddHs() and
EmbedMolecule:
supp = Chem.SDMolSupplier('4aminoacids.separated.sdf')
for m in supp:
... Chem.FindMolChiralCenters(m)
...
[(2, 'S')]
[(1, 'S')]
[(3, 'S')]
[(1, 'S')]
for m in supp:
... mH = Chem.AddHs(m)
... Chem.FindMolChiralCenters(mH)
...
[(2, 'R')]
[(1, 'R')]
[(3, 'S')]
[(1, 'R')]
Thanks for sending the file. The problem with these structures and
AddHs() was in fact the same one leading to bug 2762917:
https://sourceforge.net/tracker/?func=detail&aid=2762917&group_id=160139&atid=814650
I just checked in a fix for the bug. Your test molecules now run
correctly, as do the 4500+ in the pubchem test suite I put together.
-greg
