Two months ago I was invited by Timo Hannay of Nature to a Nature/O’Reilly FooCamp at GooglePlex. Unfortunately I was already booked and Peter Corbett was able to step in. But there was a generic invitation from Leslie Hawthorn (who has just been running the Google Summer of Code) so last week in California I took a day off the ACS meeting to go to Google and offered to give a talk about the potential of Google in Chemistry using InChI.
There are millions of known chemical compounds and they are all distinct. It’s very value to give each a unique identifier and until recently this had to be done by an authority (Chemical Abstracts Service, Beilstein, etc.). This is problematic as the numbers are copyright and you have to pay to lookup the formal link between number and compound. Recently the International Union for Pure and Applied Chemistry has developed an identifier InChI that can be automatically generated from the chemical structure with a free OpenSource program. This means that anyone can generate an InChI and the result for a given molecule will always be the same. So if we want to search for a molecule, all we have to do is generate it’s InChI and see if Google has indexed.
Yong Zhang in our group set up a server and we were able to show that they could be discovered in Google. Nick Day last year showed that they worked incredibly well. The University of Southampton crystallographer Simon Coles had put 100 compounds on the web and used Nick’s approach to add InChIs. When Nick searched for them using Google he found all 100 and no junk. This must be one of the most accurate searches ever done!
So I was able to present these ideas to Leslie and colleagues and she offered to record this on video (ca 55 min) – and they do not retain copyright. Unfortunately when I came to the GoogleInChI demo I disovered that our service had died. A pity, but all the other demos worked.
It was really nice to meet Leslie and colleagues and start planning joint activities. There is a very different attitude to that in many otehr companies. They are keen on Open Source and also looking to provide new services in GoogleBase – perhaps more of that later.
Of course Google is a commercial organisation and not a charity but there is a lot of shared vision – we have different things to contribute to the vision. For example at the eScience meeting I’ve just been at there were many demos including GoogleMaps. It has made a considerable impression.
Who knows – Googlechem?

This entry was posted in chemistry, general. Bookmark the permalink.