Seeing Genomes - Life Out of Sequence - page 165

etal changes already wrought by computers . . . computers with a win-
dow open up a whole new kind of communication between man and
machine.” 17 This last was perhaps an extreme view, but the late 1980s
and early 1990s saw the creation and use of a range of new techniques
that used visualization not just as a mode of communication, but as a
method of calculation and analysis.
Already in the 1960s, those using computers to do biological work
were using visual metaphors and pictorial means to arrange and un-
derstand their data. As I argued in chapter 5, Dayhoff's aim was the
organization of an otherwise vast and unordered glut of information
about protein sequence; the Atlas helped her achieve this end. Com-
parison of sequences can also be understood as a problem of the proper
ordering and arrangement of data (the set of all amino acid to amino
acid comparisons); this too could be understood as a visual problem
of “alignment.” These spatial organizations provided ways of ordering
biological objects that suggested meaningful connections between them.
They were not supposed to faithfully depict sequence or structure, and
they were not directed at fabrication, but rather at depicting new and
unexpected relationships between biological objects.
From AceDB to the Genome Browsers
By the early 1990s, genomic data were beginning to proliferate, and
biologists worried about their ability to synthesize and analyze all this
sequence. In 1989, Richard Durbin, working at the Laboratory of Mo-
lecular Biology at Cambridge, UK, and Jean Thierry-Mieg, from the
Centre National de la Recherche Scientifi que (CNRS) at Montpellier,
France, were trying to fi nd a way to work with data from the C. elegans
genome project. They required a system that would both maintain the
data and make them accessible to the C. elegans research community.
The physical clone map for C. elegans had been compiled using a pro-
gram called CONTIG9; this program was distributed, together with
the map data, as a program called PMAP. 18 Durbin and Thierry-Mieg
built “A C. elegans Data Base” (AceDB), as it came to be known, as an
extension of PMAP: the idea was to store sequence, physical and genetic
maps, and references within a single database. To build in the necessary
fl exibility, the authors quickly made AceDB into more than a database.
First, it became a general database management system. “Since this is
a new type of system,” the authors reasoned, “it seems very desirable
to have a database whose structure can evolve as experience is gained.
However, this is in general very diffi cult with existing database systems,