Biology Reference
In-Depth Information
affinity), potential secondary structures, cross-hybridization and probe placement;
probe design software programs such as OligoWiz (
Nielsen
et al.
,2003
) take these
parameters into account. The number of probes is obviously determined by the appli-
cation (expression microarray vs. genomic tiling array), the size of the genome and
the array formats provided by the manufacturer. For example, a
B. subtilis
genomic
tiling array has been designed using OligoWiz 2.0 (
Wernersson and Nielsen, 2005
)
and consists of isothermal probes (45-65 nt in length) starting every 22 nt on both
strands. The design contains a total of 385,000 probes matching the NimbleGen
385K array format (
Rasmussen
et al.
, 2009
).
For the design of tiling array probes, the web-based OligoWiz program, which
was initially developed for expression microarray probe selection, has been success-
fully employed (
Rasmussen
et al.
, 2009; Thomassen
et al.
, 2009
). Recently, chipD
has been created for the design of high-density tiling array probes; the program can
account for specific applications, protocols or array formats (
Dufour
et al.
, 2010
) and
has been used to design bacterial tiling arrays (
Peters
et al.
, 2009
). Another web tool
for probe selection for prokaryotic tiling arrays has been developed by
H
vik and
Chen (2010)
. A number of software programs available for the design of oligonucle-
otide probes for expression and tiling arrays have been compared and evaluated by
Lemoine
et al.
(2009)
.
In addition to array design, the specific equipment required for the hybridization,
washing and scanning of the array slides has to be considered. Microarray processing
instruments are produced by several companies, and their compatibility with the
array fabrication technology, array format and processing protocols needs to be con-
firmed. If the purchase of costly instruments or the training of personnel is not eco-
nomically justified (e.g. for a small number of experiments), certified service
providers for each of the common platforms can provide suitable, high-quality
alternatives to in-house sample processing. The Agilent Technologies and Roche
NimbleGen platforms are most relevant with respect to bacterial expression micro-
arrays as well as genomic tiling arrays, and these will be specifically addressed in this
chapter. In addition to high-density arrays, both companies offer a complete work-
flow for DNA microarray processing, including a hybridization oven or hybridiza-
tion station, an optional slide washing station and a high-resolution microarray
scanner with a pixel resolution of 2
m. These instruments can also handle slides
from other platforms, whereas the Affymetrix and Illumina systems are so-called
closed platforms, intended for the exclusive use with the corresponding arrays. It
is also possible to implement semi-automated solutions as, for example, the HS
Pro hybridization stations from Tecan for hybridization and washing. Specific pro-
tocols for one- and two-colour gene expression analysis involving these systems are
provided by Agilent.
m
2.1.3
Software requirements
The most popular software for microarray data analysis in the computational biology
community is R (
http://www.r-project.org/
) that provides a free environment for sta-
tistical computing and graphics with precompiled versions available for Windows,
