Bob,
With 12.1.4_dev the SMILES and compare issues have been pretty much
resolved:
1) A Jmol stereo SMILES submitted as input to the NIH translate site
now returns a molfile identical to the submitted model - i.e. no more
inversion.
2) smilesFromJmolMod1.find('SMILES',smilesFromJmolMod2) works in
comparing two Jmol models.
3) Using the find approach to match a Jmol model's SMILES to a JME
SMILES in an array database has one issue, but it's not really a
problem. It appears that Jmol is creating Kekule aromatic SMILES - JME
creates non-Kekule. Actually, JME may have a Kekule option - I'm not
sure. The find approach does not seem to work with Kekule finding
non-Kekule. This is no problem. There's no reason to use the JME SMILES
in the database. We can use Jmol SMILES.
There is, however, a curious point on this latter item. The strange
compare statement:
var x = compare({*},{*},'SMILES','" + pInfo1[0] + "','stddev');if (x !=
'NAN')
does make the match for aromatics! In the above, pInfo1[0] is the JME
SMILES. I'm not sure I know what the above statement is really saying. I
just made it up as a variation of one of your functions because it
seemed to be saying match this single model to a given SMILES string.
Anyway, it looks like Tom and I are back on track. I've converted our
comparisons to finds. Using Kekule aromatic SMILES in our data base is
not a problem.
Thanks for the help.
Otis
--
Otis Rothenberger
chemagic.com
------------------------------------------------------------------------------
This SF.net email is sponsored by
Make an app they can't live without
Enter the BlackBerry Developer Challenge
http://p.sf.net/sfu/RIM-dev2dev
_______________________________________________
Jmol-users mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/jmol-users
