>We have done a simple test: see the molecule called "Thiele's Ester" in the
>ECTOC Hyperglossary Tester:
>
>http://www.ch.ic.ac.uk/ectoc/glossary/
>
>This molecule is the TS of the Diels-Alder reaction between two substituted
>ciclopentadienes. As a result of the "missing" bonds, the 2D picture shows
>two issolated molecules.
>
>Now the question is: how can we improve the scripts in order to "recognize"
>transition state molecules?
This is purely a problem of using "pdb" files, which originated for
proteins and was designed around 1972. I guess transition states
were not close to the hearts of protein crystallographers! If you
look carefully, it does not even handle hydrogen atoms very well;
fairly important for organic chemists!
More modern formats (CXF, MIF) which are not yet in common
use are one solution to such problems. With Peter Murray-Rust, we are working
on an alternative known as "chemical markup language" in which
information on properties such as partial bond orders is properly
structured and included. Such methods will no doubt be a standard
component of future hyperglossaries
Dr Henry Rzepa, Department of Chemistry, Imperial College, LONDON SW7 2AY;
rzepa@ic.ac.uk via Eudora 2.1.2; Tel (44) 171 594 5774; Fax: (44) 171 594
5804.
World-Wide Web URL: http://www.ch.ic.ac.uk/rzepa.html