Biology Reference
In-Depth Information
both relational, such as SyBase and Oracle, and object-oriented, such as
ObjectStore.”
19
It was not only the quantity of the new genomic data,
but also the need for integration of different types of data that would
be required to understand them, that motivated Durbin and Thierry-
Mieg's work. A database was necessary not merely as a storage mecha-
nism, but also as a way to help the brain of the biologist to organize
and synthesize data into forms that could be used to answer biological
questions.
20
It was a tool for organizing and manipulating data in order
to help biologists fi nd interesting relationships within those data. How
could this be achieved?
Durbin and Thierry-Mieg's second innovation was making AceDB
into a visualization system. Their description is worth quoting at
length:
Overall the program is very graphical. It works using a win-
dowing system, and presents data in different types of window
according to the different types of map. The maps and other
windows are linked in a hypertext fashion, so that clicking on
an object will display further information about that object
in the appropriate sort of window. For example, clicking on a
chromosome displays its genetic map; clicking on a gene in the
genetic map displays text information about the gene's pheno-
type, references etc.; clicking on a clone displays the physical
map around it; clicking on a sequence starts the sequence dis-
play facility.
21
AceDB allowed researchers to navigate around a genome just as they
could move through the pages of the new World Wide Web. Figure 6.2
shows a screenshot of AceDB with a genome map displayed vertically
down the screen. Durbin and Thierry-Mieg realized that solving bio-
logical problems in the coming genomic age meant solving problems
of data synthesis by fi nding appropriate modes of representation and
visualization.
The designers of AceDB attempted to minimize the difference be-
tween the internal (that is, inside the computer) representation of ge-
nomic objects and their visual representations on the screen. To achieve
this, they organized the database as a series of “objects,” each having a
treelike structure; the ends of the branches of each tree pointed either to
another object or to the data themselves (numbers or characters). For
example, a “Gene” object could include branches that pointed to infor-
mation about its allele, its clone name, its physical map location, and so
