Biology Reference
In-Depth Information
facilitate the mapping of transcription units, the detection of alternative promoters
used under different experimental conditions, and the quantitation of transcript abun-
dance. Lastly, tiling arrays can be used for the genome-wide mapping of protein-
DNA interactions by chromatin immunoprecipitation (ChIP)-on-chip analysis
(reviewed by
Buck and Lieb, 2004
). On the other hand, classical microarrays
(so-called gene expression arrays), which are based on an existing genome annota-
tion and contain relatively few probes for each gene, are well suited to assess the
expression levels of annotated transcripts under many different conditions. They
are less expensive and the data easier to analyse than genomic tiling arrays.
In this chapter, we give a brief overview of microarray platform and design
options. Further, we describe well-established workflows for bacterial transcriptome
analysis including the preparation of high-quality RNA samples, cDNA labelling and
array processing. Finally,
Section 4
describes and discusses succinctly the steps that
lead to the biological interpretation of the expression profiles: from the normaliza-
tion and probe aggregation methods to the differential expression assessment and
gene regulatory networks analyses.
2
PRIOR CONSIDERATIONS
2.1
Technical requirements for microarray experiments
2.1.1
Microarray platforms
With respect to the array fabrication methods, the microarray field was initially dom-
inated by two major technologies, namely, spotted array systems and Affymetrix
GeneChip technology (photolithographic
in situ
synthesis of short oligonucleotides).
Spotted arrays are produced by depositing pre-synthesized oligonucleotides (also
cDNA or PCR products) on a glass slide. During the past decade, important techno-
logical developments in the field of
in situ
synthesized arrays, introduced, in
particular, by NimbleGen (Maskless Array Synthesizer technology), Agilent Tech-
nologies (SurePrint inkjet technology), and Illumina (BeadArray technology), have
greatly improved the sensitivity (use of longer oligonucleotides), design flexibility
(maskless probe synthesis), reproducibility, and cost-effectiveness of microarray
experiments. Importantly, higher probe densities of
in situ
synthesized arrays have
enabled tiling designs covering both strands of a given genome by overlapping
probes. A single slide is sufficient to tile a bacterial genome at high resolution, that
is, with a tiling step of less than 10 nucleotides (nt).
Because of the reproducibility of the manufacturing process, which reduces the
variability between individual slides,
in situ
synthesized arrays, unlike spotted
arrays, are not restricted to two-colour competitive hybridizations where sample
and control (or common reference) cDNAs are labelled with different fluorescent
dyes (commonly, Cy5 and Cy3) and hybridized to the same array. The relative abun-
dance of a transcript in the two samples is determined by the Cy5/Cy3 signal ratio of
the corresponding spot. The two-colour approach can reliably measure the difference
