Depth-First: Making the Case: OpenSMILES

Making the Case: OpenSMILES

Rich Apodaca — Wed, 14 Nov 2007 09:42:00 -0500

SMILES is one of the most widely-used line notations in cheminformatics. Yet until very recently, there has been no concerted attempt to develop open SMILES encoding standards.

OpenSMILES aims to change that. By providing a forum in which concerns from the SMILES user community can be voiced, peer-reviewed, and addressed, OpenSMILES introduces a new way for the SMILES language to become better.

A draft OpenSMILES specification is now available for review. For now, the best way to raise issues and otherwise get involved is through the OpenSMILES mailing list.

"Making the Case: OpenSMILES" by Craig James

Fri, 16 Nov 2007 09:45:56 -0500

Indeed, Dave Weininger always encouraged the widespread adoption of SMILES, and helped anyone who wanted to write a parser.

In my book, there are two equal halves to the definition of "open": First, the specification is available to anyone to use (Daylight's meets that criterion). And second, all experts in the field can participate on an equal footing in defining and improving the specification; the standard isn't controlled by one person or company. In other word, "no gatekeeper" as Rich says.

SMILES is a great invention, a truly brilliant solution to a hard problem in chemistry. It is simple, elegant, powerful, and easily learned. By getting the community's participation and expertise, we've given SMILES new momentum and life, solved a few of the lingering ambiguities about the language, but most importantly, we've made it a fully open specification.

"Making the Case: OpenSMILES" by Rich Apodaca

Thu, 15 Nov 2007 19:47:05 -0500

True enough. If ever there was a word that could cause confusion, it's "open." I wish I had a better, more descriptive one to use.

What I meant by "open" was firstly, a spec for which either no intellectual property, even copyright, was asserted - or which those rights were licensed in a way consistent with free use (free as in speech).

And secondly, a spec backed by a community review process where peers would work together to resolve conflicts and establish new directions. In other words... no gatekeeper.

Just my own $0.02, of course. Come to think of it, I'm not really sure there's been any discussion of what the "Open" in "OpenSMILES" means...

"Making the Case: OpenSMILES" by Andrew Dalke

Thu, 15 Nov 2007 17:58:48 -0500

To be fair, SMILES was pretty well defined in Weininger's JCICS 1988 paper, and he wrote an updated version of it for the "Handbook of Chemoinformatics" (ed. Gastinger, pub. Wiley). By most definitions that means it's "open."

However, they aren't formal in a CS sense, so there's no direct way to parse using standard tools, there are some ambiguities in a few parts, especially regarding aromaticity.