I program in python (and use NumPy, matplotlib, biopython etc.) and I think it is great. And if you use a modern GNU/Linux distribution with a large package database - like Ubuntu/Debian or Gentoo you don't even need this kitchen sink package. Then the scientific python packages are just one command away. Eg. "apt-get install python-numpy python-biopython python-matplotlib" - Anders Norgaard
I am pretty convinced by now (based on other anecdotal evidence) that Perl's days as being the main bioinformatics language are numbered. - Deepak
@Deepak Really? I'm a Python evangelist and biased, but even so, I do not see Perl prevalence decreasing in bioinformatics. For example, the traffic to the Bioperl mailing list is twice that of the Biopython mailing list. And I wonder if Jason Stajich or someone with OBF can provide an estimated download count. Perl has simply snowballed as the de facto language of bioinformatics and despite my abhorrence for it, some really useful, powerful libraries have been written in it that I wish were in Python. - Chris Lasher
@Chris But what is the rate of new development? - Kevin D. White
@Kevin I don't know. That's a very good question. What metrics could we use to assess this? I can't come up with an apples-to-apples comparison at the moment. - Chris Lasher
Perl in general seems to generate less buzz these days (e.g. see http://www.google.com/trends?q...). But if "amount of buzz generated" is a predictor of future performance, the smart money would be on Ruby... Would be interesting to do a survey of what languages are used at major bioinformatics centers (and elsewhere)? - Eric Jain
@Chris, there is still Perl code around, but a good chunk of it seems like legacy code. A lot of the companies and dev groups that I have been around are using less and less Perl (more Python, some Ruby) - Deepak
Download counts are notoriously hard to measure since one can get code from SVN or as a package from CPAN or from obf site. Website gets about 1M hits a month FWIW. I don't think Biopython should be used as a measure for python in bioinformatics though - it seems like lots of ppl are rolling their own in python and maybe that's the problem? - Jason Stajich
@Jason re: rolling their own. Absolutely correct. This is a major issue that was discussed without resolution a year ago at SciPy 2007. It led to the creation of the Biology in Python list and group. http://bio.scipy.org/ There's been chatter, but it still doesn't feel like a unification of Python efforts in bioinformatics. - Chris Lasher
I don't know about all this Ruby and Python nonsense... but *real* scientific software is still written in Fortran. ;-) - Adam Kraut
