The chemical reaction database

05 August 2012 - Cheminformatics

Some time ago Derek was kind enough to present a list of public domain chemical databases here, but what about public domain chemical reaction databases?. The inventory.

First up is webreactions, a searchable organic reaction database maintained by openmolecules.org, a group of chemical software developers. A Basic Ullmann-condensation - type reaction delivers (java applet) a fantastic 22 records. Many of them from a database called Chemsynth. This database is maintained by InfoChem GmbH part of chemReact which is part of SPRESI which appears to be started from a 1970's USSR relic. Only considers yields over 50%, so not negative results. much bibliographic info, of course redundant in 2012 with the introduction of doi's. There is a MDL number and and a RXNTYPE (info)

Spresi.com itself is linked to Springer and offers a one-week trial so we at least can have a look. The Ullmann result is 844! Additional search parameters apart from substructure search are yield range and conditions range (+800 different toluene conditions?). Peculiar: temperature, solvent and catalysts are lumped together and not addressable individually.

Back to open-source, the venerable orgsyn has a free-of-charge Chemdraw plugin for a substructure search. Searches frequently end in MolServer error 800a03ec C:\Inetpub\wwwroot\ChemOffice\CFWTEMP\OrgSyn\orgsynTemp\Sessiondir\640967013\ReactionsStructure.cdx so clearly something goes wrong (Go for Linux NOT Windows!).

Chemspider associated with the Royal Society of Chemistry also maintains a chemical reaction database. It can be searched by author, reaction type or compound. Searching for Ullmann reactions, phenol gets a hit but a substructure search on a phenol and a arylbromide results in nothing

Extending the search to the outer reaches of internet the University of Auckland promises a reaction database here but after the registration process is done it appears you have enrolled as a student. Help? Fees?. Reaccs maintained by MDL Information Systems offers an intriguing black screen, but how does it work?

And then there are the big commercial players: of course there is scifinder maintained by the American Chemical Society. Do they still charge 20 dollars per query?. If Google would get away with that sort of revenues they would own the planet by now and not just the US. Reaxys is owned by the evil empire Elsevier which we boycott.

Sord.nl offers more chemical database content based on so- called "lost chemistry" based on dissertations etc. Does it include otherwise brilliant research by people forced out of a PhD by incompetent University managers?

Late entries: Chematica @ The Guardian