Biology Reference
In-Depth Information
to the following erroneous interpretations of DNA microarray data:
(i) When mRNA levels increase, it is interpreted as an indication of in-
creased rates of transcription of the corresponding genes;
(ii) When mRNA levels undergo no change, it is taken as the evidence for
unchanged transcription rates; and
(iii) When mRNA levels decrease, it is interpreted as an indication for de-
creased transcription rates.
For convenience, we will refer to this way of interpreting mRNA levels as the
1-to-
1 interpretation
. Based on this approach, it has been widely assumed that, when
a
mRNA cluster
was found by various clustering techniques, this fact can be used
to infer that the underlying genes are transcribed with similar rates and hence that
there exists a corresponding
gene cluster
. It is one of the main objectives of this
chapter to demonstrate that this way of interpreting mRNA clusters is theoretically
invalid and factually unsupported, leading to Type I and Type II errors. A Type I
error (or a false positive) can arise, for example, when an increase in mRNA level
is interpreted as indicating an increase in the associated transcription rate (which
can be true sometimes but not always), since the level of a mRNA molecule can
increase even if the associated transcription rate does not change as long as the
number of mRNA molecules synthesized during the time period of observation
is greater than the number of mRNA molecules degraded during the same time
period. Similarly, a Type II error (or a false negative) can arise if a gene is inferred
to undergo no change in its transcription rate based on the fact that its mRNA
level did not change. This is because, even if a mRNA level did not change, the
transcription rate could have increased (or decreased) if the changes in the rate of
mRNA degradation happened to exactly counterbalance the effect of an increased
transcription rate.
Direct experimental evidence for the concept that mRNA levels, also known as
transcript levels (
TL
), is determined by a dynamic balance between transcription
and transcript degradation have been obtained only recently when
TL
and tran-
scription rates (
TR
) were measured simultaneously from human lung carcinoma
cells [8], tobacco plant cells [19] and the budding yeast
Sacharomyces cerevisiae
(S. cerevisiae)
subjected to glucose-galactose shift [10]. Here we present the re-
sults of a genome-wide analysis of the yeast
TL
and
TR
data reported in Garcia-
Martinez
et al.
[10] based on a kinetic equation relating
TL
to
TR
and transcription
degradation rates (
TD
), demonstrating the following points:
•
TL
and
TR
increase together (see Mechanism 2 in Table 12.1 and
Fig. 12.3) in 51% of the time and decrease together (see Mechanism 6)
40% of the time.
Search WWH ::
Custom Search