Graph Theory in Computational Biology: Modeling and Analysis

Abstract

This paper focuses on the use of graph theory
to model protein-protein interaction and gene regulatory
networks with the view of elucidating aspects of such
systems. Based on the STRING and KEGG databases,
5000 proteins and 25,000 interactions were included in
the PPI network analysis. Significant results are
described as follows: The degree centrality of Protein A
is the highest in the network equals to 150; the
betweenness is also high for Protein C equals to 0.072.
Several Objectives of the Analysis: There were five
functional communities identified and Community C1
associated with cell cycle and DNA repair. In the GRN
consisting of 500 genes and 2000 interactions, Gene Y
dwarfed the other genes in the network and was ranked
with the highest degree of centrality (55) and
betweenness centrality (0.085). The evaluation of
shortest path estimated the binary distance where
significant cores were recognized for formulating the
network; for instance, Path P1 = 3 and contains Genes
G001, G002, and G005. These results thus substantiate
the use of graph-based methods towards clarifying
significant proteins and control processes. As a result of
this study, it is shown how graph theory has the promise
in explaining the structural and functional properties of
biological networks and as such, provides directions that
can inform future research and development of
computational biology