Biology Reference
In-Depth Information
In my interview with one biologist who had spent signifi cant amounts
of time in industry, he spoke at length about how he understood the dif-
ference between different sorts of workers:
IT people are really like a dime a dozen . . . they are interchange-
able parts for most large corporations; but a stats guy you just
can't do that with . . . statistics is hard, computers are not . . .
I experienced this i rsthand at the Whitehead. . . . There would
be these guys that would get hired in and they would get fi red
within a month because they were just programmers. And they
would call themselves bioinformaticians. So we started to make
a distinction at the Whitehead between bioinformaticians and
computational biologists. . . . If you went to the Whitehead
or the Broad today and said: “I'm a computational biologist,”
you'd have to prove yourself. And to prove yourself you'd be
writing programs to do linear regressions models or peak fi nd-
ing for something, and if you're a bioinformatician you'd be
running the Oracle database. It's like saying “I'm a computer
scientist” versus a “I'm a computer engineer.” [At the Broad]
we were having a problem where we would hire people who
said they were bioinformaticians and we would ask them to do
something statistical or mathematical and they would just fall
on their faces. . . . We would ask them to tell us what a Z-score
means and they would be like: “uh, duh, I don't know.” How
do you call yourself a bioinformatician then? “Well I know how
to run a database and query GenBank and get some sequences
out.” Yeah, everybody on the planet can do that! We would be
like, “What do you mean, like type the sequence into the com-
mand line and get the fi le back and parse it with a parser you
wrote ten years ago?” Yeah. Oh, that's a thing people do? . . .
Millennium guys would be like “I'm a bioinformatician.” Okay,
Millennium Guy, what did you do? “I ran all the sequencing
pipelines to look at transcripts.” “Okay, cool, so did you use
hidden Markov models or Bayesian analysis to do your cluster-
ing?” And they'd be like “Huh?” So we called bioinformaticians
something different than industry was. And then the buckets
became very clear: people who didn't have math or statistics
backgrounds, those were the bioinformaticians; people who had
math and statistics backgrounds, who were so important to the
organization, those were your computational biologists. What
