Biology Reference
In-Depth Information
Dayhoff's principal innovation was not the collection of the se-
quences, but the use of this collection to investigate biology without do-
ing lab experiments. Because the
Atlas
was largely distributed on paper,
this type of investigation was at fi rst mostly limited to the NBRF. As
GenBank developed mechanisms for electronic distribution (via mag-
netic tape and over telephone-based networks), such practices spread.
GenBank
Like Dayhoff's work, the early history of GenBank must be embed-
ded within a culture of practice—databases were developed not just as
collections or repositories of data, but as tools for performing specifi c
kinds of biological work. In other words, they were active sites for the
development of biological knowledge. An account of the events that
led to the creation of GenBank has been given by Temple Smith, who
was closely involved with the events he describes.
30
Smith ascribes the
advent of sequence databases to the coincidentally simultaneous inven-
tion of techniques for sequencing DNA and of mini- and bench-top
computers. Although he describes some of the problems encountered by
the early databases, he emphasizes that the founders “foresaw both the
future needs and the potential of databases.”
31
The advocates of GenBank certainly saw the value of creating a
repository for nucleotide sequences in order to manage the output of
large-scale sequencing efforts, but they had to do much work to con-
vince potential funders and other biologists of its value. Those actively
managing the databases had to make the case that they were far more
than collections; they argued that databases should be dynamic struc-
tures and tools through which a new kind of biology could be practiced.
To most biologists, a database meant little more than an archive, not
an important tool for basic research. The caution with which the NIH
approached databases led to the construction of a “fl at-fi le” structure
for early versions of GenBank. Even this fl at-fi le database, however, had
important consequences for how biologists were able to construe and
construct the relationships between biological entities.
In addition to Dayhoff's efforts at the NBRF, several other biologi-
cal database efforts had been inaugurated by the late 1970s. In 1973,
protein X-ray crystallographic data collected by Helen Berman, Olga
Kennard, Walter Hamilton, and Edgar Meyer had been made avail-
able through Brookhaven National Laboratory under the direction of
Thomas Koetzle.
32
The following year, Elvin Kabat, an immunologist at
Columbia University, made available a collection of “proteins of immu-
