IUBio Biosequences .. Software .. Molbio soft .. Network News .. FTP

Want to move into computation biology. Advice??

Kevin Karplus karplus at cheep.cse.ucsc.edu
Mon Apr 11 11:04:54 EST 2005

On 2005-04-11, st_ev_fe at hotmail.com <st_ev_fe at hotmail.com> wrote:
> I am an experienced software developer, who wants to move into
> computational biology.
> I have no biology background, and a huge software development
> background.
> One example: I've written software which is able to do fuzzy DNA
> matching across a whole array of DNA strings, and do this in a very
> efficient time. (Usually about one millisecond even when doing a
> levenshtein operation across hundreds of thousands of DNA strings.)
> That is pretty fast, considering that the order of complexity is N^3. I
> acheived this by using an algorithm that I invented, which organises
> the DNA into trees.
> The thing is, the DNA was actually just randomly generated, as I don't
> have any idea of a real biology usage for such an algorithm. One real
> world example for this algorithm outside of biology though, would be
> for "spell checking". (There are more accurate spell checkers than
> mine, mind you.)
> Any advice where I should start? I would like to find a nice online
> community which can give me ideas and tell me needs, so that I can
> provide high performance solutions for them.

Programmers who can do high-speed algorithms are in demand in
bioinformaics, but the model you propose may not work well---the
hardest part is not usually the algorithms, but defining the problem
in a computational way.  Thus pure software developers and pure
biologists often have difficulty making headway in bioinformatics.

Your best bet is to immerse yourself in some biology, learn what data
are being collected and made public, and learn what sort of questions
one wants to ask of the data.  Developing better tools for asking
current questions and developing tools to ask new questions then
becomes more feasible.

Picking up some machine-learning methods and Bayesian statistics might
also be valuable.

A lot of the work these days is in providing easier-to-use interfaces
for biologists who do not have the time to learn complicated user
interfaces. (That's not a major focus of our work at UCSC---we're more
into the fundamental techniques and applying them to gain insight into
the biology---packaging for ease of use tends to come somewhat lower
on our priority list.)

Kevin Karplus 	karplus at soe.ucsc.edu	http://www.soe.ucsc.edu/~karplus
Professor of Biomolecular Engineering, University of California, Santa Cruz
Undergraduate and Graduate Director, Bioinformatics
(Senior member, IEEE)	(Board of Directors, ISCB)
Affiliations for identification only.

More information about the Comp-bio mailing list

Send comments to us at biosci-help [At] net.bio.net