Getting off the ground with bioinformatics?

November 10, 2010 12:54 PM

I'm a final year biochemistry undergrad, and through my internships / reading of papers / etc, I've become keenly aware that I'd like to be more conversant with computer science approaches to the field. The sheer volume of approaches and strategies in bioinformatics is a little overwhelming, though: where's a good place to start focusing my attention? I have strong maths skills, but my programming ability is limited to pecking out simple instruction sets in Python or Matlab, and my research work thus far has mostly been on isolating and characterising protein markers from serum / urine / etc.
posted by nicolas léonard sadi carnot to Education (3 answers total)

I don't have that many concrete resources to direct you to, but as a biology professor, I think that getting more versed in computational approaches is a fantastic idea. I often complain that undergraduate biology majors aren't required to take a programming course.

Two places you might start looking at: Biopython and a good matlab bioengineering tutorial from Caltech.

One final thing to bear in mind - there's a lot more you can do computationally than just bioinformatics - image analysis is another huge area, as is modeling of biochemical reactions, lots of routine data analysis, etc.
posted by pombe at 1:46 PM on November 10, 2010 [1 favorite]

First off, as pombe said, you are wise to start digging into computational tools and techniques. A pioneering computational biologist -- Lincoln Stein -- predicted that by 2012 biology and bioinformatics would be the same thing. He's off by a few years, but the spirit of that prediction is coming true in that a modern biomedical researcher needs computer skills up through basic scripting to be successful.

You mentioned having strong maths skills - extend that to statistics. Much of bioinformatics, especially in genomics, is really applied stats. A solid grasp of stats will take you far in biology, especially since that's something many of your fellow biologists will lack.

Bioinformatics is a very broad field, so it might help to narrow down what part you're really interested in. Genomics? Proteomics? Population genetics and systematics?

As for picking a language, Perl and Python are widely used in the field with strong community support - pick one of them and get familiar with it. Knowing how to do basic data manipulation with R wouldn't hurt, either.
posted by penguinicity at 4:14 PM on November 10, 2010

Speaking as a computer scientist who's dared to read biologist written perl, please choose Python. The most important part of scientific code isn't whether it fits into one line or two, but whether other people can agree that the program calculates what it sets out to.
posted by pwnguin at 5:48 PM on November 10, 2010