The CDK customizes quite a few things in the build process. One aspect of that is custom JavaDoc tags, such as @cdk.githash (source of the Taglet). This tag replaced a similar tag for Subversion (@cdk.svnrev) and allowed a link to the matching source code for that class, for which we have not found another way to achieve that. This linking functionality was broken for a while, but is now fixed again:

The last line also shows the branch name now, instead of always master, thanx to GitHub's link-to-friendly URIs for Git repository content.

Additionally, not all classes have this tag yet, and I have created a Junior Job for that.

Because I had two people asking about it, I decided to write up some material on the CDK API for handling reactions. Here's a very brief preview:

The reaction equation, BTW, was created with the mhchem package for LaTeX, which comes with Debian with the texlive-science package, which I found thanx to this Writing Chemistry with LaTeX series.
3

Computer-aided code evaluation (CACE) is an important part of scientific code development projects. There are many ways to do peer-review of source code (Maven, Gerrit, ...), and I won't go into details here. Instead, I focus on CDK's Nightly build system.

Nightly reports

Making sure the source code compiles is one of the most basic requirements.
2

I have uploaded a new revision of my Groovy Cheminformatics book, based on the CDK 1.4.11 and CDK-JChemPaint 26. Slowly I am becoming confident in uploaded PDFs to Lulu.com and perhaps the frequency of updates will increase. At least, I would love to have revisions of the book at least follow the stable releases, but the previous book version was already based on 1.4.7.

This is something I have been asked about many times. I had to find out myself, as I had no experience with this corner of the CDK rendering stack. In fact, I think there will be a second, follow-up post on that later, where I will explain I did it all wrong :)

Anyway, here is example code for how to mark a substructure. It a variations of the triazole examples I have given earlier.
1

OK, the advantage of Linked Data is that it is Linked Data. So, when a link is made to, for example, side effects, as reported below by the Free University of Berlin (using SIDER, doi:10.1038/msb.2009.98), we do not just get a link to a new resource, but we can actually look up the label for that resource, and show that in the Isbjørn results instead of the URL:

Of course, we also do have the link, so notice the link icons behind the side effect names.

Still on a tight schedule, and you must be getting tired of my updates, I'm still beefing up Isbjørn a bit more. First, I added DBPedia and FreeBase support, which means, it knows about the ontologies they use. But I also played with inline images and set the encoding so that the page not only looks nice in Bioclipse, but you can also email it and it will still look nice in Chrome and Firefox:

For FreeBase is looks similar.

Wrapping up the first release of  Isbjørn I am adding further data extraction from the databases, such as for Bio2RDF (doi:10.1016/j.jbi.2008.03.004).

The great thing about Open Standards is that they are easy to support.

I have added support for Nico's ChemAxiom to Isbjørn, here in action for ChemSpider:
Text
Text
This blog deals with chemblaics in the broader sense. Chemblaics (pronounced chem-bla-ics) is the science that uses computers to solve problems in chemistry, biochemistry and related fields. The big difference between chemblaics and areas such as chem(o)?informatics, chemometrics, computational chemistry, etc, is that chemblaics only uses open source software, open data, and open standards, making experimental results reproducible and validatable. And this is a big difference!
About Me
About Me
Popular Posts
Popular Posts
Pageviews past week
Pageviews past week
1831
Blog Archive
Blog Archive
Labels
Labels
Loading
Dynamic Views theme. Powered by Blogger. Report Abuse.