Biology Reference
In-Depth Information
was Davies, “DNA Data Deluge.” During the time of my fi eldwork, the most
attention was given to the Solexa-Illumina (San Diego, CA) and 454 Life
Sciences (Branford, CT) GS FLX sequencing machines. Solexa machines can
produce up to 6 billion bases per two-day run (www.illumina.com), while
454 claims 100 million bases per seven-hour run (www.454.com). The Broad
Institute was also beginning to test prototypes of the new SOLiD machines
from Applied Biosystems and the “open source” machines designed by George
Church's lab at Harvard Medical School. In addition, Helicos Biosciences
(Kendall Square, Cambridge, MA) was preparing to launch an even more
powerful machine capable of sequencing from a single DNA molecule (Harris
et al., “Single Molecule DNA Sequencing”).
17. During the time I spent in the lab at MIT, Burge was in the process of
trying to convince the MIT biology department to purchase a Solexa machine
to be shared between several labs.
18. A detailed review of the alternative splicing literature can be found in
Black, “Mechanisms of Alternative Pre-messenger RNA Splicing.”
19. For examples of work that asks such questions, see Wang et al.,
“Alternative Isoform Regulation,” and Sandberg et al., “Proliferating Cells.”
In the classical molecular genetics approach, the regulation of a specifi c gene
is studied, and the fi ndings from that case may or may not generalize to other
genes. The computational approach necessarily picks out those aspects of
regulation that are universal, or at least relatively common. A clear example
is the study of microRNA targets. Early studies of microRNAs focused on a
couple of specifi c examples:
lin-4
and
let-7
. These microRNAs often had target
sites that lacked perfect complementarity to the 5
end of the microRNA (the
“seed”), obscuring the importance of this region. This problem was quickly
corrected by computational studies. See Lewis et al., “Prediction of Mamma-
lian MicroRNA Targets,” and Rajewsky, “MicroRNA Target Prediction.”
20. For example, see Fischer, “Towards Quantitative Biology.” This review
gives examples of how bioinformatic knowledge is expected to be able to
generate quantitative predictions about diseases and interaction of drugs with
the body.
21. Fortun, “Projecting Speed Genomics.” “Speed genomics” refers to biol-
ogy done on a large and expensive scale, in big, centralized labs, with expen-
sive machines.
22. Fortun, “Projecting Speed Genomics,” 30.
23. See chapter 3 for a more detailed discussion of notions of the regimes
of “productivity” associated with computing.
24. The chimpanzee paper is fairly typical of the kind of large-scale work
that the Broad undertakes. Its website lists active areas of research as “deci-
phering all the information encoded in the human genome; understanding
human genetic variation and its role in disease; compiling a complete molecu-
lar description of human cancers . . .” In addition to many genome projects,
its work has included attempts to completely characterize human genetic
variation (HapMap, 1000 Genomes), to completely characterize cancer and its
genetics, and to create an RNAi library that covers every known human gene.
′
