in reply to Other Bioinformatics Monks Out There?

Some define bioinformatics narrowly as computational genomics and proteomics. Others use bioinformatics more broadly as a term for computational biology in beneral.

I fall into the latter category. I am a computational neurobiologist who studies how neurons in sensory systems code information using action potentials or "spikes". The analysis is purely mathematical and draws from algorithms in signal processing, nonlinear physics, and information theory.

Perl is my first choice of language for programming in this domain. It allows me to try ideas quickly and is often fast enough for what I need. Some algorithms are CPU and/or memory intensive, in which case I'll turn to the Perl Data Language (PDL) to cut the problem down to size. PDL is the tool of choice for signal filtering, FFTs, and correlational analysis.

I occasionally write applications in C++, but it is more to keep up my skill set than desperate need. Others in my department use Perl or Python for a similar tasks. I think Python seems to attract more numerical analysts for some reason, but PDL is nicer than the melange of numerical modules one has to deal with in Python.

I have also worked in bioinformatics in the narrow sense. I used Perl to attack the protein folding problem using a genetic algorithm Monte Carlo technique. I am interested in doing more work in genomics and proteomics. There are a lot of fascinating problems in those domains.


  • Comment on Re: Other Bioinformatics Monks Out There?

Replies are listed 'Best First'.
Re^2: Other Bioinformatics Monks Out There?
by BioGeek (Hermit) on Aug 03, 2004 at 21:50 UTC
    And what is your background? Are you a neurobiologist by training who happend to play a lot with programming, and found that he could put that to use in his work? Or are you a programmer by trade who realised that some of the more interesting problems - with real-life applications - are to be found in biology?
      I am a theoretical physicist who decided there is a lot of potential for quantitative techniques to be applied to neurobiological problems.

      Theoretical physics is often portrayed as one of the more difficult, deep sciences. While it is true that physicists often use esoteric mathematics, physicists pick their problems and theories to be analytically tractable, or at least analytically approximable. Biology is none of this. Biological systems are nonlinear, nonequilibrium, and nonstationary. Dealing with such systems is at the forefront of physical techniques and beyond. That is where the fun is.

      I've had a love of programming and hacking ever since I was a kid, so it is natural for me to attack these problems computationally. Perl came to the fore mostly because I can program in perl faster, with fewer bugs, than any other language I have learned before or since. In addition to PDL mentioned before, perl is used for keeping track of experimental databases, data munging, GUI design through perl /Tk, and glue that holds the whole data analysis protocol together.