Biology Reference
In-Depth Information
biological objects involve decisions about what features of an object are
represented and how they are related to one another.
When an object is represented in an image, it is poured into a care-
fully contrived structure. As with databases, decisions about visual-
ization techniques and tools are decisions about how to constitute an
object. These decisions, in turn, affect how the biological object itself
can be understood. Bioinformatic “pictures” do not attempt to capture
what life looks like, but rather to fi gure out how it works. The most
important visuals, the ones that bioinformaticians use every day, are
not like images from a microscope, or even like particle tracks compu-
tationally reconstructed from a spark chamber, which are both traces
of physical entities. Rather, they break traces into pieces, reorder them
with computations, and reconstruct them into pictures that are easily
interpretable and full of biological meaning. These pictures are usually
the starting points, rather than the end results, of bioinformatic work.
These pictures are what biology looks like to many bioinformaticians—
they are the raw material with which experiments can be performed.
Visualization in bioinformatics is—through and through—analysis and
quantifi cation.
2
As the amount of biological data, especially sequence data, has in-
creased, the usefulness of thinking about biology merely as a long series
of AGTC words has diminished. In 2012, GenBank contained around
150 billion nucleotides—this is an inconceivably long “book of life”
(more than fi fty thousand times the length of the Bible). Ways of render-
ing data into images, rather than text, have gained in importance. Doing
things with large amounts of biological data requires thinking of them
as a picture, not merely as a text. The transition from text to image is an
ongoing process, but historians and anthropologists of biology should
be attentive to the new metaphors—visual and pictorial ones—that will
increasingly dominate the discipline.
3
The Ensembl database, at the European Bioinformatics Institute, is
a sophisticated example of a “genome browser,” an online tool that bi-
ologists can use to see and study genomes. This representation of the
genome is almost entirely a visual one; through it, the user-biologist
“sees” the genome as a set of color-coded horizontal lines. Of course,
this representation is not intended to be a faithful depiction of the intri-
cate, tangled-up set of molecules of a real genome. But in an important
sense, this is what the genome is—this is how it looks and feels—to
many biologists; visualization techniques become the only way of seeing
the object in question.
4
Large matrices or grids are also important for
thinking about problems in bioinformatics. Reducing multidimensional
