[Biopython-dev] GenePop parser code

Peter (BioPython Dev) biopython-dev at maubp.freeserve.co.uk
Wed Dec 27 18:40:03 UTC 2006


Tiago Antão wrote:
> Hi!
> 
> I have finalized converting my code to parse GenePop(*) files into
> BioPython. I would like to know if there would be interest in adding
> it to BioPython?
 >
 > (*)GenePop: http://wbiomed.curtin.edu.au/genepop/
 >

I think someone asked about Population genetics on a mailing list in the 
last year or so, and there was some interest out there.

Check Feb 2006 on the discussion list, look for a thread on "PopGen" 
started by Ralph Haygood.  He had written a Python module for computing 
a variety of population-genetic statistics from DNA sequences.

Does this overlap with your plans at all?

> After reading the instructions, I suppose, as this is not strictly a
> patch, that it makes no sense to submit it through Bugzilla? More, as
> this is my first contribution to BioPython, I would bet its full of
> small problems typical from a newbie...

I would suggest creating an enhancement bug for "Adding a GenPop file 
format parser" or something like that, and then attach your code to it. 
  Sending attachments on mailing lists is asking for trouble.

> I also have test code, but its still in my own ad-hoc format (I will
> convert it to the BioPython standard in case it gets accepted).

Have a look at the existing Tests directory - basically you have a 
script that produces some output, and the expected output in the 
"output" subdirectory.  Avoid things like printing floating point 
numbers or filenames in any print statements - they make running the 
test suite on different platforms tricky.  Most parser tests have a few 
sample input files too (held in subdirectories again).

> I have truckloads more of Population Genetics code that I would be
> willing to convert to Biopython (more or less things along the lines
> of BioPerls's Bio::PopGen). I decided to start with a small/simple
> thing to see how it goes...

That sounds sensible - unless there is some generic object structure you 
want to introduce for handling population data?  I don't know enough 
about this to comment.

> Is there any interest? If so, how do I submit this?

To recap: yes there is some interest, and I would suggest creating an 
enhancement "bug" on bugzilla and attaching files/patches there.

Peter



More information about the Biopython-dev mailing list