Bioinformatics degree questions.
oggied at hotmail.com
Fri Apr 5 21:51:22 EST 2002
On Fri, 29 Mar 2002 20:16:16 -0600, deodiaus wrote:
> I was thinking of going back to school, and getting a Ph. D. in
> Bioinformatics and trying to land a job in industry. However, I am 40,
> and maybe a bit too old for this approach (even going back to do a M.
> Sc. program in BioInfomatics). I do have a M. Sc. in Applied Math, and
> have been programming for 15 years (http://deodiaus.tripod.com).
> Instead, I was giving thought to maybe going back to do a Ph. D. in Math
> and specializing in a research topic in Bioinformatics. Unfortunately,
> getting through the Math prelims and other exams is a major obstacle.
> The other problem is that many math Ph. D. programs are quite
> theoretical, and not always pragmatic. I love the theory, however, a
> lot of employeers will probably look at me and think this guy will just
> not work out, he is too theoretical, and not able to solve the problems
> at hand in the time frame that we are looking at (2 years). On the other
> hand, I have to say that "good" programming is an art, and takes years
> to accomplish in order to master correctly. I was giving thought to
> looking (and rearchitecting) a program like (JHU's) George Rose's
> program LINUS, which attempts to model protein folding. Maybe, the LINUS
> needs a good rewrite by a seasoned programmer. Currently, I am suffering
> from bad design decisions, and think a rearchitecture will solve lots of
> problems, so maybe this might be an approach to Linus. All the physics
> can be right (but I don't think they are), but bugs may exist in code
> which are eternally hidden in the woodwork.
I dont know about you but I dont think writing software is that of a good
venue to prove your skills in the academic field. I am coming out of a
BSc in computer science in Canada and I got a summer job with the National Research
Council of Canada (even got rehired after that) where I rewrote clustalw to be
multi-threaded and work
on virtually any posix compliant unix (even cygwin would do with proper
pthread support). Yes, there is clustalw-multithreaded from SGI but they
wont give you the source code and it only runs on SGI IRIX. The version I
wrote will be available shortly to the general public (within a month)
together with source code as an open source project. However, I found
that it really doesnt matter when looking for jobs or trying to get into
MSc programs. What academy and employers want are good or excellent
grades and they are stuck to it. Unfortunately my grades suck... :)
My 2 cents.
More information about the Comp-bio