Immediate justification for cloud-based cheminformatics apps

Last week I found myself in New Jersey participating in a discussion between TM Forum and Pistoia Alliance attendees about a topic of interest to many informatics scientists – how can we go from creating a mobile App store to a cloud-based life science R&D information ecosystem. While I am probably limited in what I can say my experiences during and after travel made it pretty clear why this needs to happen sooner rather than later.

I gave a presentation on Apps for rare and neglected diseases, which admittedly is limited to a few so far. But a certainly think it could be a growth area. I described Open Drug Discovery Teams and TB Mobile, both developed with Alex Clark (who was also at the meeting). One of the ways I found to justify the TB Mobile app surprised me in response to a question I was asked (why use a mobile app and not another computer?). I thought briefly of a use case, say I attend a TB conference and I see someone give a talk and they show the structure of a molecule, with my mobile app I can quickly sketch the molecule and do a similarity search and find the closest structure and its potential targets, pathway etc. I can then ask the author a better, more informed question..your molecule looks like X, have you tried target Y or looked at pathway Z? As someone who tends to thrive on the spontaneity of conferences and find new collaborators through such interactions, this is just one real-world example I could see of how such apps could be transformative. Now imagine if I was able to run the molecules I see presented in a talk or poster against computational models or dock them via apps that run the jobs on the cloud, then near instantly return the results. This is pretty close to doable now. It does not feel like its forcing the issue. The right people to make this happen were in the room. I hope this blog adds a little urgency.

During my travel, my expensive laptop (4 month old Sony) crashed twice. Today it failed for good. While I back up my files, some of the cheminformatics software I use almost daily requires considerable time to download and set up. I am pretty constantly creating 3D conformer libraries so I can perform searches of vendor compounds for ordering. Needless to say I doubt I will be doing this on a temporary machine (used to write this blog). So I will wait until I have a functioning hard drive again. This hammered home another use case for having apps that use the cloud, in particular cheminformatics apps, because then I could submit jobs from my iPad or iPhone and not have to rely on software residing on a dead PC. While I guess working from a laptop has been fine for a number of years, doing the types of repeated chminformatics tasks is losing its allure. What if I could generate my databases for searching and have them on the cloud forever and point to them when I need to do a search? Similarly what if all the things I do (and publish on for target X, transporter Y and receptor Z) could all be done on the cloud via mobile app. The following couple of slides were used at the end of the talk as a suggestion or wish list of what I would like developed – call this a manifesto of sorts or call to arms for App development. My function could be replaced by a few apps. Yes I realize that, but it could also free me up to work on other tasks. Bizarrely, these slides were sketched several weeks ago on my iPad, little did I know that they would be so urgently needed with the passing of yet another PC.

One other thing I picked up on from the meeting was how attendees at the meeting had missed a relevant source of app information for drug discovery in scimobileapps.com, fortunately Antony Williams was also at the meeting to talk about the Chemspider app and also highlighted scimobileapps. It will be interesting to see how the cheminformatics ecosystem evolves and this website may be a good way to track its growth (e.g. take a look now compared with a few years from now), and I have a very real vested interest in making it happen.


Leave a Reply

Your email address will not be published.

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <s> <strike> <strong>