Johannes joined a Bioclipse Workshop a long time ago, and introduced the participants to the idea of using XMPP (aka Jabber) for asynchronous web services. SOAP is commonly user to run webservices over HTTP, but via (SMTP) email and XMPP is possible too (see SOAP over XMPP). Using HTTP as transport layer has problems. The biggest problem, is possibly that HTTP connections are timed out, e.g. by intermediate router.
After some difficulties this week with making an export of CDK plugins in the Bioclipse2 Cheminformatics feature of with the cdk-eclipse software, I got the following cute Bioclipse2 script up and running:
Recently, I have been blogging about Git:
A new environment means new tools. Bioclipse is Eclipse RCP-based, so colleagues work with Eclipse and are much more into Eclipse too. For example, into Mylyn. Mylyn is a tool to track tasks and assign context to them. The tasks I am interested in (for this blog item), is fixing bug reports. Mylyn is rather suited for this, as it allows linking Java source files to bug reports.
Five days ago, my chem-bla-ics blog turned 3. Here’s the first post . It defined:
About a week ago, I hooked up my GitToDo software with Freemind. This allows me to organize the projects I am working on, without having to code this in GitToDo directly.
As you might have seen, we, Uppala and the EBI, are working on the next generation JChemPaint. JChemPaint is an editor, and therefore, consists of a model (IChemModel), a view (IRenderer) and a controller (IController). See the many posts in Gilleain’s blog.
Some time ago Noel ran a poll on chemoinformatics and cheminformatics , so I set up a poll too in part #1 of this series . The outcome is clear: The Obernai meeting strongly suggested chemoinformatics [1], but the start of the open access Journal of Cheminformatics is the killer.
Just updated Bioclipse2 with Jmol 11.6 RC 18:
Lee et al. published last week a paper on pKa prediction (doi:10.1021/ci8001815). As the paper says, the pKa, and in particular the ionic state of a molecule at physiological pH, affects pharmacokinetics and pharmacodynamics. The paper describes a (binary) decision tree using presence or absence of SMARTS substructures to traverse the tree, allowing prediction of monoprotic molecules.
There was some talk about the history of chemoinformatics toolkits by Noel and Andrew, which made me wonder on the exact history of Jmol and JChemPaint.