Biology Reference
In-Depth Information
25. The Solexa machines allowed the direct sequencing of mRNA taken
from cells. Older methods usually required mRNA to be reverse-transcribed
into DNA before sequencing. This was usually done in short fragments known
as expressed sequence tags (ESTs).
26. Previous studies include Nigro et al., “Scrambled Exons”; Cocque-
relle et al., “Splicing with Inverted Order of Exons”; Capel et al., “Circular
Transcripts”; Zaphiropoulos, “Circular RNAs”; Caudevilla, “Natural Trans-
splicing”; Frantz et al., “Exon Repetition in mRNA”; Rigatti et al., “Exon
Repetition.”
27. One bioinformatic study of exon scrambling had been performed using
ESTs, concluding that scrambling events were either nonexistent or extremely
rare: Shao et al., “Bioinformatic Analysis of Exon Repetition.”
28. Compare my description with Galison's description of the discovery of
neutral currents in the 1970s: Galison,
How Experiments End
, 188-198.
29. Fieldwork notes, August 9, 2007.
30. Of course, it is literally possible to go through a program line by line to
understand its behavior. However, this is almost never how it is done. Rather,
you debug a program by running it on some real data, seeing what happens,
seeing whether the results make sense. If not, you use the error to fi nd the part
of the program that needs attention, alter it, and try again.
31. In my work, the Solexa machines themselves were regions of
ignorance—for proprietary reasons, Illumina did not reveal all the details
of how their sequencing process worked, and hence the characteristics and
reliability of the data it produced were largely unknown. The lab did exten-
sive computational analysis on the data it received to try to fi nd the extent to
which it could rely on those data.
32. See Popper,
Logic of Scientifi c Discovery
.
33. Bacon,
New Organon
, 110.
34. On “discovery science,” see Aebersold et al., “Equipping Scientists,”
and Boogerd et al.,
Systems Biology
.
35. Allen, “Bioinformatics and Discovery.” For further discussion of
hypothesis-driven research, see Wiley, “Hypothesis-Free?”; Gillies, “Popper
and Computer Induction”; Allen, “Hypothesis, Induction-Driven Research and
Background Knowledge”; Smalheiser, “Informatics and Hypothesis-Driven
Research.”
36. This debate has been carried on vigorously in the blogosphere. See, for
instance, Young, “Hypothesis-Free Research?”
37. “Proposals to base everything on the genome sequence by annotating
it with additional data will only increase its opacity.” Brenner, “Sequence and
Consequences.”
38. Allen, “In Silico Veritas,” 542.
39. In later chapters, we will detail more precisely the methods used for
data management and data analysis in bioinformatics.
40. Garcia-Sancho, in particular, has argued for the importance of spe-
cifi c computing architectures in shifting the practices and epistemologies
of nematode worm biology (see Garcia-Sancho, “From the Genetic to the
