Biology Reference
In-Depth Information
network can be analysed, at least in part, by breaking it down into motifs whose
behaviour is simple and well understood.
A number of motifs have been demonstrated to be statistically over-represented
in metabolic and transcriptional networks. Motifs such as small feed-forward loops,
single input modules and cycles have been identified as over-represented compared
with equivalent randomly connected networks in
E. coli
(
Shen-Orr
et al.
, 2002;
Dobrin
et al.
, 2004
) and
S. cerevisiae
(
Wuchty
et al.
, 2003
). It is often assumed that
the motifs are over-represented because of positive selection pressure. However,
Mazurie
et al.
(2005
) compared over-represented motifs in the transcriptional
network of
S. cerevisiae
with those in a number of other hemiascomycetes, and
concluded that the regulatory processes for the biological function under consider-
ation were dependent upon post-transcriptional regulatory mechanisms rather
than transcriptional regulation by network motifs. They concluded that the presence
of motifs is unlikely to provide a selective advantage to the organism, possibly
because they are deeply embedded in the rest of a complex network of genetic inter-
actions. Despite this controversy, the interactomes of a wide range of microbial
genomes have been searched for network motifs in an attempt to understand the
genome-wide, systems-level dynamics of the functional networks (
Herrg
˚
rd,
2004; Stelling, 2004; Gelfand, 2006; Janga and Collado-Vides, 2007; Ravcheev
et al.
, 2011
)
Software Availability
Cytoscape:
http://www.cytoscape.org/
.
Ondex:
http://www.ondex.org/
.
STRING:
http://string-db.org/
.
GeneMania:
http://www.genemania.org
(Yeast only).
BioPixie:
http://pixie.princeton.edu/pixie/
(Yeast only).
6
THE ROLE OF eSCIENCE
One of the most promising approaches to tackling very large data sets is that of
global, collaborative analysis. Exchanging data and results across social, political
and technological boundaries is a challenging process, both socially and technolog-
ically. Over the last decade there has been increasing research into, and use of,
eScience. The term eScience refers to collaborative, global science that is performed
in silico
with a computational infrastructure (
Luciano and Stevens, 2007
). The major
umbrella technologies supporting eScience are Grid and Cloud computing.
Grid computing was named by analogy to the electricity grid. The aim of Grid
computing is that plugging in to worldwide compute resources should be as easy
as accessing electricity (
Kesselman and Foster, 1998
). Grids are characterised by
their geographic distribution: datasets and compute resources are held in different
