Biology Reference
In-Depth Information
correct, extend, or otherwise relate to previous work, and as a
result a substantial number—in fact, a majority—of database
entries have to be updated each year; thus GenBank is not an
archival operation such that an entry, once made, just stays in
place.
54
By this time, some members of GenBank's scientifi c advisory panel con-
sidered the growing backlog an “emergency.”
55
Los Alamos responded
by requesting more money to employ more “curators” to enter data
from the published literature. Goad and his co-workers, however, real-
ized that the root of the problem was that the structure of the database
was increasingly inadequate for the needs of biological research. James
Fickett and Christian Burks, leading contributors to the Los Alamos ef-
fort, argued that “the scope and interconnectedness of the data will grow
at a pace hard to keep abreast of,” and that consequently, the greatest
challenge would be to “organize the data in a connected way.”
56
Because the NIH saw the nucleotide sequence database as a mere
archiving activity, they attempted to create an atheoretical database.
This was impossible: even the minimalist fl at fi le encoded a particular
structure, a particular way of doing biology, and a particular idea about
how sequences related to organismic function. The fl at-fi le structure in-
stantiated an ordering of biological elements based on the one gene-one
enzyme hypothesis. During the early 1980s, that hypothesis was in the
process of being displaced and superseded by other ideas about how
biology worked.
Biological Relations
The original GenBank contract ran for fi ve years, expiring in September
1987. As that date approached, two concerns were paramount. First,
GenBank continued to struggle to remain up to date in entering se-
quence information from journals.
57
Second, it was clear that the struc-
ture of the database required a signifi cant overhaul. As such, NIH's new
request for proposals specifi ed that the contractor for the next fi ve-year
period would develop a new system whereby authors would be able
to submit their sequence data directly in electronic form (preferably
over a dial-up telephone network). In addition, the contractor would
be obligated to fi nd ways to increase the cross-referencing of the data
and to make sure that “new data items which become important can
be added to the data base without restructuring.”
58
The NIH received
three “competitive” proposals for the new contract: one from BBN, one
