Thursday, December 30, 2010

Computational Biology and Pseudoscience?

"If I were starting today" as a scientist in high school or college, “I’d be in computational biology” - Francis Collins
This has been my favorite quotation for sometimes now. I sometimes hear people talking about Bioinformatics as pseudoscience and now I begin to wonder is that really so? I am working from home, and thus raising very many eyebrows about the "quality of science" I many be doing. For records, I was de-selected for a job that once I thought I was selected just because I work from home. Well, this types of prejudices do exist and is certain to be there, but we may need to learn to live through it and fight it when possible.

I was examining what a wet lab scientist does vis a vis a computational scientist. Remember computational scientist does not have to step outside of the comfort of her home to do her job.
A wet lab scientist reads paper, plans experiment and goes to the lab. Prepares solutions, buffers, and sets up an experiment. It may be running a PCR, or making a strain grow or checking the gels etc.. The computational scientist also reads a paper (often hard hitting papers with complex mathematical formulae and algorithms), plans to work on the data and run them. Running experiments for a computational scientist would mean running some applications(mostly freewares that comes without any warranty or without any documentation). This process may take anytime between hours to days. In the end, the computational scientist sits with the output and checks if anything worthwhile has come out of it. Often this is a very laborious process. It may so happen that your desirable output is not there after you run it for days!! Then you may need to switch to another algorithm or change parameters(like repeating experiment with changed growth condition or a changed buffer). If things work the way you wanted, you may call the experiment was successful. After some "Successfully" run experiments, the scientist sits back and stitches all the outputs together to form a story. If publishable - hurray you got a paper. And the same happens with the wet lab researcher albeit the process may be slightly more manual intensive and slow. So, both the processes are quite comparable and parallel to each other. So, why call computational science as pseudoscience??

No comments: