Biology Reference
In-Depth Information
The complexity of this system results from the fact that huge amounts
of data need to be made accessible in fractions of a second. Storing the
data, accessing the data, and displaying the data each have their own
appropriate modes of representation. Doing bioinformatics means solv-
ing the practical problems of how to transform and manipulate these
representations of biological objects into forms that can be used by
biologists. Understood in another way, HTML, SQL, C, and so forth
embody distinct logical domains in which objects must be structured
according to particular rules; particular representations are ways of
translating or mediating between those domains. Programs like autoSql
generate boundary objects that can be passed back and forth.
But bioinformaticians solving such problems also face another set of
even more important issues. Each of the representations of the genome
must not only be interchangeable with the others, but also make sense of
the genome as an object in the real world. Brian Cantwell Smith makes
the point that the “choices one makes about the
content
of the various
pieces of a representational system—how the representation represents
its subject matter as being—are often affected by how one describes the
computational structures themselves, the structures that carry or have
content.”
38
Decisions about how to represent an object entail decisions
about what that object must be in the real world. Biological objects
like genomes are not something that we can pick up and hold in our
hands—we do not have an intuitive ontological grasp of what they are.
Computational representations of a genome—for instance, what is part
of the genome and what is an extraneous “label”—describe what a ge-
nome is.
The UCSC Genome Browser interface itself consists of a series of
“tracks” displayed one below the other horizontally across the browser
window (fi gure 6.4). The horizontal distance across the screen corre-
sponds to the position in the genome, such that reading vertically down
a column allows comparison of different features at the same genomic
position. “The graphic display of the browser,” the creators write, “is
invaluable for getting a quick overview of a particular region in the
genome and for visually correlating various types of features.”
39
The
tracks represent genes (as discovered by various methods), exons, gaps,
expressed sequence tag alignments, cross-species homologies, single nu-
cleotide polymorphisms, sequence tagged sites, and transposon repeats.
Users can also add their own “custom tracks” to the image, which al-
lows them to view, for example, how their own positioning of genes lines
up with the standard annotation. Clicking on particular tracks leads
to more detailed information: links to other databases or information
