Polymers are where the problem with assuming that chemistry is structure centric come home to roost of course. Nico Adams has been looking at developing ways of describing polymers effectively and it really isn't straightforward. The whole notion of substance, chemical, and structure get very very messy. - Cameron Neylon

What is distressing is that someone added that initial obviously incorrect SMILES and it propagated easily to all those databases. As Tony has said many times curation is a big challenge in chemical databases. - Jean-Claude Bradley

I'd say the cause of everyone blindly copying data, is that they are happy enough to be able to copy data at all... it all starts with Open Data and the right to fix things, and share those fixes... - Egon Willighagen

+1 Egon - Cameron Neylon

I'm not sure that the reason people are copying data are because it's Open. PubChem continues to proliferate yet it is NOT Open. We've had this exchange before..PubChem is NOT Open data but I "judge" it gets copied because people treat it as an authority. It is not a good idea to treat PubCHem as an authority as it is a repository...non-curated and with no efforts underway to curate it that I am aware of. This does NOT mean that it is not useful...just that users must beware and caution is required.. That said, I agree with the right to fix things ...ChemSpider can be fixed... - Antony Williams

I agree that PubChem should not be treated as authoritative, but the big advantage it has over ChemSpider is that it can be downloaded and re-used. - Michael Kuhn

I agree that people copying PubChem is indeed not because it is Open... that one is copied merely because it is free and confuse that with Open. That said, I love to see the day that there was a court ruling on that state of the PubChem data, which is at some places claimed to be Public Domain, something refering to the copyright being with the providers, while at other places... a bullet proof statement would be *very* useful indeed. - Egon Willighagen

Michael - there is every intention to provide access to the ChemSpider structure collection in the near future. This will not include all associated information in a record as the associated information has mixed licensing but free to use. - Antony Williams

ChemSpider and Wikipedia have the advantage that errors can be corrected. Are there mechanisms in place to correct errors in DrugBank, PubChem and the many other derivative databases? - Jean-Claude Bradley

DrugBank: yes, David Wishart is said to be quite responsive to comments. Same for KEGG. PubChem: NO. They just aggregate tons of source databases, and you would have to hunt down the source and hope it makes it way downstream... Thus ChemSpider's annotation efforts are invaluable, and I would love to be able to re-use them - Michael Kuhn

Michael...drop me an email at antonyDOTwilliamsATchemspiderDOTcom and let's discuss how you want to use information and what you need to get at and it might be an issue of simply pointing you to the right web services. Lots of people are using ChemSPider through web services at present. We'd love to help you - Antony Williams

thanks Michael - Jean-Claude Bradley

There is also a curated KEGG: http://biometa.cmbi.ru.nl/ - Egon Willighagen

Biometa appears to be static since 2007. The SDF file mostly contains R groups. Seems weird - Antony Williams

Martin Ott moved to Sheffield... (http://www.lhasalimited.org/index...) I'll try to ping Vriend about the future of BioMeta... or the code behind it. - Egon Willighagen