Computational Systems Biology Methods and Protocols.7z

connected to everyone else through a very short path. And there

are some other parameters to measure the structure of the network

including the network diameter, the betweenness centrality, the

average path length, the closeness centrality, and the network den-

sity. Network diameter measures the network size which is the

length (in number of edges) of the longest geodesic path between

any two vertices. The betweenness centrality is a measure of cen-

trality in a graph based on shortest paths, and a node with higher

betweenness centrality would have more control over the network.

The closeness centrality is a measure of centrality in a network,

calculated as the sum of the length of the shortest paths between

the node and all other nodes in the graph. Thus the more central a

node is, the closer it is to all other nodes. The network density is the

ratio expressing the number of actual edges between vertices to the

number of possible edges.

2.2.2 Differential
Coexpression Gene Sets
(Module) Identification

Numerous methods and tools were developed to measure the

differential coexpression of genes including methods to identify

differentially coexpressed gene modules which can be gene sets

newly detected [9, 16–18] or a set of genes predefined [10, 19,

20 ] (Table2). WGCNA is the most commonly used tool to detect

the coexpressed modules. It took the soft-thresholding strategy in

constructing the coexpression network and defined a dissimilarity

measure for gene nodes. Then the average linkage hierarchical

clustering coupled with the dissimilarity was applied to identify

the coexpression modules. Then a preservation analysis can be

applied to test whether the modules detected in one condition

preserved in another condition. For the method coXpress, genes

are clustered by the Pearson correlation coefficient (r), and the

gene-to-gene distance can be calculated by 1r. Then the resulting

tree will be cut by a predefined cutoff value getting a series of gene

Table 2

Methods for differential coexpression module identification

Method/tool

name

Gene sets

predefined

Extending to multiple

conditions References

WGCNA No No [9]

coXpress No No [17]

DiffCoEx No Yes [18]

DICER No No [21]

GSCA Yes Yes [19]

GSNCA Yes No [20]

CoGA Yes No [10]

160 Bao-Hong Liu