When we went live with ChemSpider in late March we took advantage of some commercial tools available from our collaborators to assist in the release. Our intention has always been to shift our dependence to Open Source tools where appropriate both to reduce our dependence on commercial tools as well as to allow us to use and enhance Open Source tools as appropriate for our needs and return our enhancements back to the community.

At present we are replacing our SMILES generation capabilities with those provided by the OpenBabel chemical toolbox. OpenBabel is well proven and has been used to support over 30 projects. In order to facilitate interaction between other databases utilizing canonical SMILES generated using OpenBabel adoption of the generating tool is the best approach of course!

We are also going to utilize the InChI DLLs directly rather than depend on a third party implementation for InChI support. Utilizing the DLLs directly will actually reduce the time for generation, very important when you generating them for 14 million compounds!

While these changes will not change the user experience our hope is for greater integration and independence! We are presently considering the addition of other Open Source components and are researching the availability of 2D structure cleaning open source algorithms. Any ideas?

Stumble it!

One Response to “Adding Open Source Tools to ChemSpider for SMILES and InChi Support. Structure Cleaning Anyone???”

  1. baoilleach says:

    Great to hear you’re using OpenBabel (I’m a dev). Regarding 2D structure cleaning, what exactly do you mean? The CDK has 2D diagram generation.

Leave a Reply