Biology Reference
In-Depth Information
is enacted. In the labs I visited, these virtual spaces were not only harder
to fi nd and harder to access, but more intricate and more diffi cult to
navigate, than the physical spaces. Indeed, a large part of learning to do
bioinformatics is the process of learning to fi nd one's way around the
appropriate virtual spaces—having the expertise to know
where
some
piece of information or code might be was often as valuable and as
valued as knowing what to do with it. This chapter draws on my experi-
ences at the Broad Institute and the European Bioinformatics Institute
(EBI) in order to describe what these virtual spaces of bioinformatics
look like, how people move around in them, and how this movement
constitutes the
doing
of bioinformatics.
The fi rst part of this chapter details how biology becomes virtual in
the fi rst place—how material samples become data as they pass through
the sequencing “pipeline.” This process is at the center of bioinformat-
ics: showing how the material becomes the virtual is to show how bi-
ology becomes digital. What becomes apparent in the description of
the pipeline, however, is that this fl attening of objects into data is a
complex and contested process—samples do not automatically render
themselves into zeroes and ones. Although a pipeline suggests linearity,
what we fi nd here is that the process is “out of sequence”—it is messy
and contingent. Following the data to their origins shows how much
their generation depends on care and judgment and how much work
must be done to make standardized, virtual biological objects that are
amenable to computers and networks.
In the second part, I examine another way in which biology is ren-
dered into data. Producing “ontologies” means creating standardized
common languages for speaking about biology. The uniformity and
universality of biological objects is manufactured during bioinformatic
work. Such standard, computable objects are necessary for bioinfor-
matic work to be possible. Bioinformatics standardizes and fl attens bio-
logical language into data. Third, I examine how bioinformatic objects
are moved around. The rendering of biology into standard, computable
formats allows knowledge to be produced by the careful movement and
arrangement of data in virtual space.
What emerges here is a new picture of biological work and bio-
logical knowledge. Biological work consists in the rendering of biologi-
cal stuff into data and the proper arrangement of these data in virtual
space. Neither the creation of data nor their movement in virtual space
is a wholly linear, fl uid, or frictionless process. Data must be carefully
formed into their digital shapes, and their arrangement must accord
with standardized structures such as standardized data formats, ontolo-
