Dalke Scientific Software: More science. Less time. Products
[ previous | newer ]     /home/writings/diary/archive/2008/11/06/teaching_python_for_cheminformatics

Teaching Python for cheminformatics

I've started doing training courses for computational chemists who want to learn more about Python. All the examples I use come from cheminformatics, using OpenEye's OEChem toolkit. If you are interested, I'll be teaching a two-day course in the Bay Area on December 4-5, a three-day course in Leipzig on 2-4 March, and I've started planning a course for Boston in early April. See my web site for more details or contact me directly.

Python is quite widely used in cheminformatics, with the major competitor languages being C, C++ and Java. I say that but they aren't really competitors because Python can work together with those languages. Most of the cheminformatics toolkits, after all, are written in one of those languages, and many have Python bindings. You can use PyBel to use OpenBabel from the C implementation of Python, or use Jython to talk to the CDK. OpenBabel even has .Net bindings that you can use from Iron Python.

Last month I did a two day course in Leipzig, hosted through Mike Müller's Python Academy. All of the examples were drawn from cheminformatics, using OEChem as the main cheminformatics toolkit. It's the toolkit I know the best, and I think it's the most powerful and flexible available. I also covered using Python to talk to a database, make plots, communicate with Excel, and other special topics.

The biggest feedback I got was that I tried to cover too much during those two days, and didn't give enough hands-on time. I've did some on-site teaching for AstraZeneca yesterday and another tomorrow, bearing that in mind. In one day I covered the basics of Python and of OEChem. I deliberately did not cover exceptions, classes, or more advanced topics because I don't think they are that important for what most computational chemists do. They use classes but rarely make their own.

My students yesterday were quite happy with the class. There's a few things to tweak - list comprehensions are too complicated for an intro course. I'll see how things go tomorrow.

If you're interested in one of my classes, or want to come to your site to do training, please contact me.


Andrew Dalke is an independent consultant focusing on software development for computational chemistry and biology. Need contract programming, help, or training? Contact me



Copyright © 2001-2020 Andrew Dalke Scientific AB