Biology Reference
In-Depth Information
ally effi cient solutions to this problem that the UCSC Genome Browser
becomes a powerful tool for providing biological insight. The solution
that the genome browser uses is primarily a pictorial one: its represen-
tation of the data as a dynamically generated set of images gives it the
ability to show that which is of biological interest. The data themselves
exist as a vast mass of hierarchical and cross-referenced tables within a
computer. What the biologist wants is specifi c information about, for in-
stance, the location or size or annotation of a gene. It is in and through
the genome browser images that these two domains interact—that data
can be extracted from the databases in ways that biologists can use.
The genome browser is a visual solution to the problem of organizing
and accessing large volumes of computational analysis in a biologically
meaningful way. Moreover, solutions to the problems of how to repre-
sent objects within the computer and on the screen generate particular
ways of seeing and interacting with biological objects; in an impor-
tant sense, these computational representations reconstitute the objects
themselves.
Many other examples could be given in which computational biolo-
gists use images to work with large sets of data or with computations.
In genomics in particular, tools such as the UCSC Genome Browser are
used to visualize genomes not just as sets of As, Gs, Ts, and Cs, but as
one-dimensional maps showing the positions of chromosomes, genes,
exons, or conservation. All these examples use machine-generated im-
ages to render the invisible visible. The representations are a way of see-
ing, organizing, and surveying large amounts of data. Their aim is both
depiction of something “out there” in the biological world and some-
thing “in there,” inside a computer. It is the representations themselves,
in the images and in the databases, that stand between these two do-
mains. These representations are valuable tools because they constitute
objects in ways that allow them to be drawn into new and nonobvious
relationships with one another.
Solving Problems with Pictures
None of these systems of representation are made up by biologists on
the spot. As these examples show, they are embedded in the histories
of software and hardware. But we also want to know how biologists
actually go about solving problems by making pictures: What do they
actually do with these images? This section describes the use of heat
maps, a new tool for visualizing alternative splicing, and the work of
a computational information designer. In these examples, reducing a
