Biology Reference
In-Depth Information
Chapter 10
Boolean Models of Cellular Signaling
Networks
Zhongyao Sun
1
and R ยด ka Albert
1
,
2
1
Department of Physics, The Pennsylvania State University, University Park, PA 16802, USA,
2
The Huck Institutes of the Life Sciences, The Pennsylvania
State University, University Park, PA 16802, USA
Chapter Outline
Introduction 197
Boolean Networks and Biological Systems 198
Boolean Network Modeling 199
Reconstruct the Network Based on Biological Knowledge 199
Determine the Boolean Transfer Functions
Model Validation: Reproduction of Known Results
205
Robustness against Disruptions and Useful Implications
206
Application Examples
207
T-LGL Leukemia Network Modeling
207
200
PathogeneImmune Response Network
207
Updating Schemes and Incorporating Time
201
Conclusions and Future Directions
207
Network Initialization
202
References
208
Steady-State Analysis
202
INTRODUCTION
Network structures permeate every sphere of cellular
biology. The vast number of intra- and extracellular
processes and interactions that form a complex web of
mass, energy and signal transfer can intrinsically be
described in network language. The nodes of a biological
network are cellular components such as genes, RNAs,
proteins and small molecules, and the edges are reactions,
interactions, and regulatory or synthetic relationships
among components. The process of transcribing coding
DNA into mRNAs is either promoted or suppressed by
transcription factors; the totality of these transcriptional
processes can be integrated into gene regulatory networks
[1]
. Similarly, diverse interactions among proteins, or the
biochemical reactions in cellular metabolism, can be
readily depicted in network language
[2
e
4]
. In a network
representation the elements of the system are represented
by network nodes, and the interactions among elements as
edges. The wealth of data and the affinity between network
science and biology make network modeling of biological
systems not only viable, but also powerful and uniquely
useful (see Chapter 9).
In addition to the network structure (that is, the nodes
and edges), it is often desired to be able to describe the
dynamics of mass or information transfer through the
network. Dynamic models characterize the nodes by states
(e.g., concentration or activity), and the states of the nodes
change in time according to the interactions encapsulated in
the network. Continuous dynamic models use sets of
differential equations to capture the detailed variation of
concentrations of key substances
[5]
in the system
(see Chapter 16). Despite the recent phenomenal growth of
computing power, such approaches become practically
impossible to implement when the number of nodes in
a system reaches more than 100. Difficulties are also posed
for parameter estimation in large-scale systems when there
are insufficient temporal data. The aim to circumvent these
difficulties makes discrete dynamic modeling such as
Boolean networks
[6
e
13]
and Petri nets
[14
e
16]
particu-
larly useful. Vast amounts of qualitative knowledge
regarding regulatory relationships between cellular
components have been either experimentally measured or
inferred from biological evidence. The compilation of such
knowledge forms the basis for network reconstruction,
which then allows discrete modeling of a system, bypassing
the obstacles posed by parameter estimation. The model
can be tested against experimental evidence and refined
iteratively. Discrete dynamic models generate insightful
