Biology Reference
In-Depth Information
FIGURE 6.6
Visualization of alternative exon splicing. This image is similar to a genome browser image,
displaying a region of sequence horizontally. In this gene, two possible splice patterns are observed,
one including exon 3A but leaving out 3B, and the other leaving out 3A but including 3B. The image
shows that the former splice pattern tends to occur in skeletal muscle and heart tissue (although not all
the time) and the latter tends to occur in testis and liver tissue (although in testes, 3A is occasionally
included). (Wang et al., “Alternative isoform regulation.” Reproduced by permission of Nature Publishing
Group.)
a large amount of meaningless irregularity. Lab members worked to
tune the image so that only the salient features would be brought into
focus.
Figure 6.6 shows an image produced by the visualization tool. The
linear layout of the exons is displayed across the bottom, while four
different tissues are displayed in different lines at the top (rendered in
different colors in the original). Most signifi cantly, arcs connect exons
to one another to show the splicing pattern. In the case of this gene
(AK074759), we can see that splicing in skeletal muscle and heart cells
predominantly connects exon 2 to exon 3A to exon 4, while splicing in
liver and testis cells predominantly makes the connection via exon 3B.
Such an image provides clear evidence that, at least for this gene, the
way in which exons are spliced could play a signifi cant role in making a
skeletal muscle cell different from a testis cell or a liver cell. It does so by
summarizing and organizing vast amounts of data and computational
work. The small graphs appearing over each exon show the number
of sequences counted beginning at each nucleotide position. Since the
scale is logarithmic, the whole image captures information gleaned from
thousands of 32-base-pair Solexa-generated sequences. These sequences
were then subjected to computational processing to map them to the
human genome and to organize them into a “digital archive.”
Coming to terms with this large, multidimensional data set meant
producing a visual system. It was not the case that the lab members
understood the role of alternative splicing in cell differentiation before
doing the work to make it visible; knowledge about this feature of biol-
ogy was produced through the development of the visualization tool.
It was in following this work that I discovered that some of the lab
members were thinking seriously about the problems that visualization
presented to biology. Indeed, Monica, a lab member working on the
