Biology Reference
In-Depth Information
detected and quanti
ed through a variety of
antibody-based methods, most of which look at
only one or a few proteins. This limitation has
recently been overcome by protein arrays
that allow for high-throughput measurements
of many proteins simultaneously, generating
a data set capable of ef
physiological conditions. Two types of data are
frequently used to construct cellular networks.
The
first is the primary data of identi
ed and
quanti
ed macromolecules (RNA, protein, or
metabolite). The second set consists of cellular
networks that describe relationships between
the identi
cation of
different groups or individuals.
1
The drawback
to the array technique is that it is a targeted
approach, detecting only proteins targeted for
measurement. The alternative is to use quantita-
tive proteomics as an unbiased method to
measure hundreds of proteins within a sample.
2
Tremendous advances in both sample prepara-
tion and mass spectrometry (MS) technologies
3
make this approach very appealing for directly
linking an underlying default in a biological
system to a functional macromolecule.
This approach
cient classi
ed components and is constructed
from functional data sets. Several types of
networks have been constructed including
gene regulatory, genetic interaction, and
protein
e
protein interaction networks. These
networks are frequently depicted in a similar
manner in which a node is the molecule that
has been identi
ed such as RNA, DNA, protein,
or metabolite. The lines, or edges, represent rela-
tionships between molecules, such as direct
interactions, enzymatic modi
cations, or tran-
scriptional regulation. Because this representa-
tion allows easy viewing of data, a quick
assessment of the relationship can be made by
generating a
fits very well with the
emerging notion that a more holistic approach
is required to comprehensively understand
complex biological systems such as cells,
tissues, or organs. This stems from the fact
that almost all functions are carried out by
complex networks and any perturbation can
lead to abnormal phenotypes that are re
figure that shows the connectivity
between the identi
ed macromolecules (
Box 1
).
Macromolecules with many connections are
often described as hubs, and many have been
shown to be key players in maintaining homeo-
stasis in biological systems.
5
Perturbation of
a single gene that encodes for a hub protein
can potentially lead to complex phenotypes
because it affects several protein
e
protein inter-
actions resulting in abnormal outputs from
several different functional pathways. For
example, it has been shown that highly
connected proteins that are de
ected
in the expression, connectivity, or function of
many components.
4
It is logical to conclude
that evaluating only one or a few components
of a system will not fully predict its behavior.
The
same argument
could be made
for
a biomarker.
Its predictive power might be
signi
cantly reduced if it is not a part of the
primary network associated with the biological
process that it is designed to evaluate but rather
a secondary consequence that might be associ-
ated with many other biological perturbations.
This setup would lead to signi
ned as hubs in
protein
e
protein interaction networks have
a higher probability of encoding disease-related
genes.
6,7
A high-quality protein
e
protein interaction
network is informative and can provide valu-
able insight
cantly reduced
speci
city for the marker, as it would not be
able to distinguish between different biological
states.
Several different cellular networks have been
constructed and regularly used in a systemwide
effort to better understand the behavior of
complex biological
into the function of
individual
proteins and the network
s biological processes.
Because it is becoming increasingly apparent
that the complex phenotype of many diseases
involves alterations of several different
proteins, a detailed mapping of their interac-
tions is essential
'
systems under different
for a better understanding
Search WWH ::
Custom Search