Thursday, August 11, 2011

CDK 1.4.1: the changes, the authors, and the reviewers

It seems I had forgotten to blog about the stable update for CDK 1.4, but here it is for 1.4.1 (download). I also finally get fed up with searching my blog each time for those git scripts for commit statistics to write up this post, so I now posted them in this gist. I also took the opportunity to now point to GitHub commit pages, rather than those on SourceForge.

As a decent stable update release, nothing much happened. A good part is atom types: I added a few myself, and the first bits of work of the patch by Nimish and Gilleain in Asad's team made it in. Nimish has done a great job over the summer to find the details (some details are still missing, such as hybridization info for many metallic elements) for atom types found in KEGG. Besides some minor code clean up, only one other thing happened. The addition of InChINumbersTool, which you can read about already in my book (see figure), but I'll be blogging about that pretty soon too.
  • Ge final f3a8828
  • K final 8bae588
  • Li final 2641484
  • Hooked in iodine atom type detection 3d4f7a5
  • I final 82a60f0
  • Na final c2554bb
  • Hooked in As atom type detection 2ed384b
  • As final f0a327d
  • Si final c3d28b6
  • Added attribution 50ef575
  • Moved countTestedAtomTypes() to the end, to fix the checking if all atom types have been tests; also added a comment that is must be the last method 229f036
  • Mn final 105850e
  • Added missing dependency a872522
  • Added a helper class to get atom number for the heavy atoms, using the InChI algorithm abcc00a
  • Added the P.ane atom type: 5-coordinate neutral P 90371fd
  • Added a sp1 phosphor type, as found in PubChem CID 138843 (reported by Julio) d0e0bf2
  • Added a charged phosphor type, as found in PubChem CID 20643237 (reported by Julio) 19e0875
  • Added missing atom type properties for Ne, Mn.2plus, Fe.2plus, Al, Ni.2plus, Mg.2plus, K.metallic, (addresses #3355759) dae49f5
  • More descriptive variable names (fixes #3310938) and full copyright statement e871b67
The Authors

12  Egon Willighagen
  8  Gilleain Torrance

Gilleain extracted the small, single atom type patches, but, as said, much of the work was done by Nimish.

The Reviewers

  6  Egon Willighagen 
  6  Rajarshi  Guha 


  1. I appologize for a offtopic.
    Is there any free alternative to chemaxon reactor? A program with ability to identify reactive groups and linking them together? Thanks.

  2. No problem :)

    The Blue Obelisk eXchange is an excellent place to ask questions like these, but it is welcome here too :)

    You could have a look at CDK-Taverna, which has a reaction enumerator. I am not really familiar with the ChemAxon products (as I still have not found the time to have their licenses checked to see if they conflict with my own open source cheminformatics research), but guess this might be close to what you are looking for.