tag:blogger.com,1999:blog-7844526396210378482.post5699088378197787442..comments2024-01-31T09:23:26.925+00:00Comments on Noel O'Blog: Cheminformatics toolkit face-off - LogP and TPSANoel O'Boylehttp://www.blogger.com/profile/03288289351940689018noreply@blogger.comBlogger3125tag:blogger.com,1999:blog-7844526396210378482.post-81588576388151779582009-03-10T09:16:00.000+00:002009-03-10T09:16:00.000+00:00Good post Noel, your keeping track of things!!!. I...Good post Noel, your keeping track of things!!!. I am reading gasteiger book (cover to cover) and developing apps using ruby.<BR/><BR/>By the way, I wanna say that<BR/>"Ruby my Love!"<BR/>"Ruby the future!!"<BR/><BR/>python is not good some how...OOPs suks badly. "___" suks too much...<BR/><BR/>Lot of syntax inconsistencies in python but lot of libraries are there.<BR/><BR/>Once Ruby has as much libraries the name "python" will be deleted from my head.<BR/><BR/>My choice "(Ruby + JRuby) ||<BR/> (Ruby + C)" <BR/>for any task.<BR/><BR/>Love,<BR/>Chemboy.Anonymousnoreply@blogger.comtag:blogger.com,1999:blog-7844526396210378482.post-49045738930258133722008-04-24T15:03:00.000+01:002008-04-24T15:03:00.000+01:00Thanks for the encouraging comments, Joerg. For su...Thanks for the encouraging comments, Joerg. For sure, there are problems at several levels. Perhaps the aim of this 'first generation' of cheminformatics toolkits is simply to identify which problems are the most serious going forward, so that subsequent versions will be built to handle such problems from the start, e.g. Rich's series on "Can your cheminformatics tool handle this?".Noel O'Boylehttps://www.blogger.com/profile/03288289351940689018noreply@blogger.comtag:blogger.com,1999:blog-7844526396210378482.post-92080044100436465992008-04-22T19:22:00.000+01:002008-04-22T19:22:00.000+01:00Noel, I really appreciate this effort and I hope t...Noel, I really appreciate this effort and I hope that you keep track of the differences ;-)<BR/><BR/>If you are unlucky the same SMARTS patterns might lead to different results, since the chemical perception cascades might be very different (aromaticity, hybridisation, kekulization, stereo assignment, and finally atom typing).<BR/><BR/>This was basically one reason why people on OpenSmiles could not easily agree on an aromatic standard. If there would exist one, we could just share the text definition for it ;-)<BR/><BR/>I said this already several times, if we want clean solutions we need text, better XML, perceptions, <B>only</B>. This would require to modularize single perception layers, which take XML input. The XML can then be distributed and all packages are fine, just by implementing a simple XML to perception layer, which would be the same ;-)<BR/><BR/>Problem, we are far from having a query language strong enough for this task. I was hoping that OpenSmiles might serve as starting point, but the discussion seems stalled. MQL is still my favourite for several reasons, but I guess I am not objective on this.<BR/><BR/>Anyway, great work ... keep going all together ... accepting the problem is already a good way in improving problems ;-)Anonymoushttps://www.blogger.com/profile/09112376168632883058noreply@blogger.com