provides the set of interacting parts undergoing transformation,
which can enhance our understanding of allostery at the fundamental level. These developments are presented in Subheading 4. A case
study on G-Protein Coupled Receptors (GPCRs) is presented in
Subheading 5, illustrating the application of these methods to
obtain better insight in allostery in these receptors.
2 Introduction to Networks
For exploring the behavior of a collection of entities, the relationship and dependencies between them should be known. For example, the behavior of a committee of people can be studied by the
knowledge of the committee members and the nature of interaction with respect to each other. These entities, along with their
inter-relations, can be represented as a network, in which the entities
are named as nodes and the connections between them are called
edges. The connections between the nodes can be binary (existing
or non-existing) or can vary along a scale of values. Depending on
the type of connections, the edges can be constructed as
unweighted (binary) or weighted. The range of the weights depends
upon the system and the type of problem that is being studied. A
network can be represented as a graph or an adjacency matrix (row
and column arrangement of numbers) of size n  n, where n is the
number of nodes in it. Once the network related to the system
under scrutiny is constructed, it can be studied through various
basic metrics such as degree, hubs, and clustering coefficient to gain a
better insight about the system. Cliques and communities represent
higher-order connectivity in a network, and they capture the local
geometries in detail, within the framework of global topology of
the network. All these parameters of a network can be enlisted and
studied using various software like GRAPROSTR [16] and CFinder
[17] by giving the adjacency matrix as input.
Though unweighted (binary) networks are easier to comprehend, the connections in most of the real-world networks (like
social networks, metabolic networks, and protein structure networks) are non-binary. It is not only whether a connection exists
between two nodes that matters, but the strength of the connection
is more important here. In this context, weighted networks play an
important role in studying real-life situations. Also, apart from the
above-mentioned parameters, weighted networks can be analyzed
by ego-net (sum of weights of all edges) of a node and many other
parameters specific to weighted networks. For example, the shortest
path between specific nodes can be evaluated using methods such as
Dijkstra’s algorithm [18], and has been adopted to PSNs [19] in
elucidating the paths of communication due to allostery. Despite
advancement in quantitatively capturing specific parameters, the
matrix of a network contains a wealth of information which is
Network Re-Wiring During Allostery and PPI
91
which can enhance our understanding of allostery at the fundamental level. These developments are presented in Subheading 4. A case
study on G-Protein Coupled Receptors (GPCRs) is presented in
Subheading 5, illustrating the application of these methods to
obtain better insight in allostery in these receptors.
2 Introduction to Networks
For exploring the behavior of a collection of entities, the relationship and dependencies between them should be known. For example, the behavior of a committee of people can be studied by the
knowledge of the committee members and the nature of interaction with respect to each other. These entities, along with their
inter-relations, can be represented as a network, in which the entities
are named as nodes and the connections between them are called
edges. The connections between the nodes can be binary (existing
or non-existing) or can vary along a scale of values. Depending on
the type of connections, the edges can be constructed as
unweighted (binary) or weighted. The range of the weights depends
upon the system and the type of problem that is being studied. A
network can be represented as a graph or an adjacency matrix (row
and column arrangement of numbers) of size n  n, where n is the
number of nodes in it. Once the network related to the system
under scrutiny is constructed, it can be studied through various
basic metrics such as degree, hubs, and clustering coefficient to gain a
better insight about the system. Cliques and communities represent
higher-order connectivity in a network, and they capture the local
geometries in detail, within the framework of global topology of
the network. All these parameters of a network can be enlisted and
studied using various software like GRAPROSTR [16] and CFinder
[17] by giving the adjacency matrix as input.
Though unweighted (binary) networks are easier to comprehend, the connections in most of the real-world networks (like
social networks, metabolic networks, and protein structure networks) are non-binary. It is not only whether a connection exists
between two nodes that matters, but the strength of the connection
is more important here. In this context, weighted networks play an
important role in studying real-life situations. Also, apart from the
above-mentioned parameters, weighted networks can be analyzed
by ego-net (sum of weights of all edges) of a node and many other
parameters specific to weighted networks. For example, the shortest
path between specific nodes can be evaluated using methods such as
Dijkstra’s algorithm [18], and has been adopted to PSNs [19] in
elucidating the paths of communication due to allostery. Despite
advancement in quantitatively capturing specific parameters, the
matrix of a network contains a wealth of information which is
Network Re-Wiring During Allostery and PPI
91