Abstract
Graph clustering algorithms are widely used in the analysis of biological networks. Extracting functional modules in protein-protein interaction (PPI) networks is one such use. Most clustering algorithms whose focuses are on finding functional modules try either to find a clique like sub networks or to grow clusters starting from vertices with high degrees as seeds. These algorithms do not make any difference between a biological network and any other networks. In the current research, we present a new procedure to find functional modules in PPI networks. Our main idea is to model a biological concept and to use this concept for finding good functional modules in PPI networks. In order to evaluate the quality of the obtained clusters, we compared the results of our algorithm with those of some other widely used clustering algorithms on three high throughput PPI networks from Sacchromyces Cerevisiae, Homo sapiens and Caenorhabditis elegans as well as on some tissue specific networks. Gene Ontology (GO) analyses were used to compare the results of different algorithms. Each algorithm's result was then compared with GO-term derived functional modules. We also analyzed the effect of using tissue specific networks on the quality of the obtained clusters. The experimental results indicate that the new algorithm outperforms most of the others, and this improvement is more significant when tissue specific networks are used.
Highlights
Graph clustering of PPI networks is one of the most common techniques for inferring functional modules [1,2,3,4,5]
There is no widely accepted formal definition of a functional module, it is commonly conceived as a group of proteins that work together to carry out a cellular process while binding to each other in different times and places [6].Various graph clustering approaches have been developed in order to discover sets of densely connected vertices within a graph
For example in our analyses QCUT and HQCUT found good resulting clusters on Homo sapiens, but they could not find any clusters on C.elegans and Yeast PPI networks
Summary
Graph clustering of PPI networks is one of the most common techniques for inferring functional modules [1,2,3,4,5]. Most clustering algorithms with the purpose of finding functional modules in bioinformatics, try to find either highly dense sub graphs based on finding cliques or in a greedy manner grow clusters starting from vertices with high degrees as seeds. CFinder finds all k-cliques that are defined as complete sub-graphs with k vertices (k$3), and it merges kcliques if they share exactly k-1 vertices The other algorithms such as SPICi [14] and clusterONE [15] optimistically suppose that the clusters in PPI networks are located around the vertices with high degrees. Modularity is a quantitative measure that was originally defined by Newman and Girvan [23], to assess the quality of graph clustering results Based on this measure, a clustering algorithm gets high modularity if it finds clusters with dense connections between the each cluster’s inside and gets low modularity scpre with sparse connections. Choose an edge with the highest weight and decides whether or not the two edge’s endpoints are in the same cluster or not
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
