Biology Reference
In-Depth Information
cultural, political, and economic consequences.
2
But as well as ordering
society, databases construct orderings of scientifi c knowledge: they are
powerful classifi cation schemes that make some information accessible
and some relationships obvious, while making other orderings and re-
lationships less natural and familiar.
3
Organizing and linking sequence
elements in databases can be understood as a way of representing the
connections between those elements in real organisms. Like a billiard-
ball model of a gas in physics, databases do not aim to be a straightfor-
ward representation of a biological system; rather, they aim to capture
only some of its important features. The database becomes a digital
idealization of a living system, emphasizing particular relationships be-
tween particular objects.
As I worked with biological databases in my i eldwork, I started to
ask why the information in them was arranged the way it was. Indeed,
how did databases become the preeminent way of storing biological
data? Answering these questions required an interrogation of the his-
tory of databases. By examining a database diachronically, we can dis-
cover how changes in structure correspond to changes in the kind of
work being performed (and in the knowledge being produced) through
databases.
The different database structures that GenBank has used represent
different ways of understanding and ordering biological knowledge.
Early “fl at-fi le” databases, such as those constructed by Margaret Day-
hoff and the fi rst iterations of GenBank, instantiated a protein-centered
view of life in which single sequence elements were placed at the center
of biological understanding. The “relational” databases that gradually
replaced the fl at fi les in the 1980s and 1990s emphasized the intercon-
nections between sequence elements—biological function was pro-
duced by interaction between different elements, and the connections
were refl ected in the database. Finally, the “federated” databases of the
postgenomic era, while still placing sequences at the center, allowed
much wider integration of other (extra-sequence) data types. This gave
structural expression to the notion that biological function could be
best understood by modeling the relationships between genes, proteins,
transcription factors, RNA, small molecules, and so on. By following
data into databases, we see how the rigid structures of information tech-
nologies impose constraints on how data can move and be shaped into
knowledge.
The activities of data storing and knowledge making are not sepa-
rate and are not separable. Biological databases are not like archives
and museums—they are oriented toward the future more than the past.
