[BioPython] "Developing Bioinformatics Computer Skills"

Brad Chapman chapmanb@arches.uga.edu
Wed, 28 Mar 2001 19:07:12 -0500


Me:
> I'm a biologist who tried C++, Java and Perl before deciding python was the
> right bet for me.

Johann tries to start a language war: 
> And I'm a computer scientist[1] who has used and continues to use a whole
> slew of programming languages, and though I have yet to find the perfect
> language for teaching programming to a complete newcomer, Perl is very high
> on my list of languages which are utterly UNsuited to that task.[2]  If you
> want to teach someone all the wrong habits, start him off on Perl.

Ewan obliges with a counter-attack:
> It depends whether you want to teach somsone programming or get something
> done (honestly). People who indulge in my-langauge-better-than-yours >
> /dev/null  in my view. Clearly are not interested in solving real
> problems.

Sorry guys -- I didn't mean at all to start a language-war-thread. I was
definately just talking about my personal situation and not trying to
make any kind of justification for python as the best thing for
everyone and every task. Python just fits well with my brain 
(gawd, I hope that's not insulting to python :-), and makes me 
enjoy programming and allows me to be productive. Hell, that's good
enough for me. Other people can program in assembly for all I care, as 
long as it makes 'em happy.

me:
> Count-me-in-for-a-Python-and-Bioinformatics-book-ly y'rs;

Dinu signs me up as an author:
> Seriously, what makes it so hard to assemble a few minds from 
> this list to spend a few weeks working on one chapter each, 
> with somebody else plus a publisher like O'Reilly doing the 
> editing? 

Seriously, I am but a humble graduate student. I'm in no position to
pretend to be qualified to write something like this. Personally, I'm
happy to have the Biopython libraries available, and have a nice bunch 
of docs to go with 'em. 

Johann lays the groundwork for a python book bigger then War and Peace:
> I've been thinking for some time that I should put
> together an "introduction to programming for bioinformatics", using Python as
> a vehicle.  In my copious free time these past six months, I've even written
> one half of one tutorial.  :-)  Unfortunately I've found that top-down,
> though great in theory, does not work that well in practice.  You sort of do
> need to know what a variable and a for loop are before you can be taught
> about classes.  Hmm.  I'm open to suggestions.  I'm also very open to the
> idea of such a course being developed communally.  :-)

Well, I don't really agree with your top-down approach. I am more of a 
practical person and would rather know how to count how many records
I've got in my FASTA file then to spend chapters reading about
people's pontifications on objects. For me, it took a long time of 
being hit over the head with stupid solutions I came up before I could 
see a good reason to obsess over how I should use objects best. Of
course, that is probably because I've got a thick skull :-).

I do think having more written material on "starting with python in
bioinformatics" would be very useful. If you wanted to expand on the
current Tutorial (big plug time!):

http://biopython.org/wiki/html/BioPython/BiopythonCode.html

I'd be happy to help with that. What is currently there assumes you
already know something about python programming, but I don't even know 
how helpful people find it. Well, anyways, I use it quite a bit when I 
forget how to do things :-).

Writing-chapter-1-of-Bioinformatics-Programming-in-Assembly-ly y'rs,
Brad