Biology Reference
In-Depth Information
managing large data sets would be rewarded both fi nancially and in
terms of prestige.
This hierarchy is built into the structure of the Broad itself, which is
divided into “programs” (such as the Cancer Program, the Psychiatric
Disease Program, and the Program in Medical and Population Genetics)
and “platforms” (such as the Genome Sequencing Platform, the Imag-
ing Platform, and the Biological Samples Platform). Broad personnel
generally understand the programs to be performing the real research
work of the institute, while the platforms provide support in the form
of imaging, cheminformatics, and sequencing. The Computational Biol-
ogy and Bioinformatics program recruits mostly individuals with strong
mathematics and statistics backgrounds, while the lower-status “bio-
informatic” work is distributed among the supporting platforms.
8
We should not be too hasty in following practitioners in dividing
bioinformatics into “higher” and “lower” forms—both contribute to
knowledge production in crucial ways. The attitude that informatics is
less valuable to an organization than computational biology is contin-
gent upon a particular notion of “value” for biology. The private sec-
tor (represented by “Millennium Guy”) holds a different view of what
count as valuable skills for biological work. For pharmaceutical compa-
nies like Millennium, value is generated by “high throughput”: fi nding
potential drug targets by testing thousands of proteins; fi nding a chemi-
cal with drug potential by screening hundreds of chemicals; or predict-
ing potential toxic effects by checking thousands of possible “off-target
effects.” Such tasks do not necessarily require a detailed understanding
of how a particular biological pathway or system works; since the space
of potential compounds and potential targets is immense, speed and
volume are crucial.
But this kind of bioinformatics is not only valuable within the pri-
vate sector: these kinds of practices are also highly regarded
within
spe-
cifi c types of biological work. For instance, in the Broad Sequencing
Center, the aim is to produce as much sequence as quickly and cheaply
as possible with the smallest number of errors. Workers who can de-
sign better databases, streamline sequencing pipelines, and fi nd ways
to process sequence data more quickly are highly sought after. Bren-
dan, a programmer, told me how he designed a Laboratory Informa-
tion Management (LIM) system—an internal lab database—that greatly
improved the sequencing process. The SQUID-BetaLIMS system, which
was still used by the Broad during my fi eldwork, allowed the lab to
track samples, diagnose errors, and track workfl ow in ways that were
crucial to its performance as a state-of-the-art sequencing laboratory.
