Biology Reference
In-Depth Information
or coordinates on a principal component axis (
Zhang
et al.
, 2008
). Another algo-
rithm, FIRE, uses clusters but allows the motifs to extend across several clusters
(
Elemento
et al.
, 2007
). Finally, we mention our own implementation of an idea
to combine expression profiles and motif discovery, HMMTREE, whose sequence
model accounts for pairwise correlations between promoter activities summarized
in a hierarchical clustering tree to discover Sigma factor binding sites (
Nicolas
et al.
, 2012
).
5
FINAL COMMENTS
The analysis of bacterial transcriptomes with expression microarrays and genomic
tiling arrays, respectively, has reached a mature stage with respect to experimental
procedures and data analysis strategies. It allows processing of large numbers of sam-
ples in a relatively short period of time and at relatively low cost compared to the
emerging RNA-seq approach. Gene expression microarrays, although based on an
existing genome annotation, are still adequate tools to accommodate systems biol-
ogy, but with obvious limitations when compared to unbiased whole-transcriptome
methods. In particular, RNA-seq technology shows great promise with single nucle-
otide resolution and higher sensitivity and dynamic range than tiling arrays, even
though it is still less well-established and lacks widely accepted standards for data
generation and analysis (
M¨der
et al.
, 2011
). Indeed, the microarray field has led
to such significant efforts towards methodological refinements over a period of
15 years that, even if the technology is finally discontinued, understanding the basic
principles that have been developed will certainly be of interest for future technol-
ogies as it is already true for RNA-seq.
References
Alter, O., Brown, P. O., and Botstein, D. (2000). Singular value decomposition for genome-
wide expression data processing and modeling.
Proc. Natl. Acad. Sci. U.S.A.
97,
10101-10106.
Aubert, J., Bar-Hen, A., Daudin, J.-J., and Robin, S. (2004). Determination of the differentially
expressed genes in microarray experiments using local FDR.
BMC Bioinformatics
5, e125.
Ackermann, M. and Strimmer, K. (2009). A general modular framework for gene set enrich-
ment analysis.
BMC Bioinformatics
10, e47.
Bailey, T. L. and Elkan, C. (1994). Fitting a mixture model by expectation maximization to
discover motifs in biopolymers.
Proc. Int. Conf. Intell. Syst. Mol. Biol.
2, 28-36.
Benjamini, Y. and Hochberg, Y. (1995). Controlling the false discovery rate: a practical and
powerful approach to multiple testing.
J. Roy. Stat. Soc. Ser. B
57, 289-300.
Blom, E. J., Ridder, A. N., Lulko, A. T., Roerdink, J. B., and Kuipers, O. P. (2011). Time-
resolved transcriptomics and bioinformatic analyses reveal intrinsic stress responses
during batch culture of
Bacillus subtilis
.
PLoS One
6, e27160.
