Protein-to-protein interactions: Technologies, databases, and algorithms
ACM Computing Surveys (CSUR)
Mining biological interaction networks using weighted quasi-bicliques
ISBRA'11 Proceedings of the 7th international conference on Bioinformatics research and applications
| Hi-index | 0.00 |
miRNAs are fragments of non-coding RNA that play a key role in mRNA regulation. Unfortunately complex relations among miRNAs and mRNAs remain still unclear, although there is a substantial agreement on the fact that miRNA are regulator of the mRNA expression. The complete elucidation of such mechanisms may have a great impact in the elucidation of cancer development and progress. In particular the comprehension of the disruption of regulatory mechanisms related to diseases remains still an un-explored field. Such relations may effectively modelled as a bipartite (un)weighted graph as proposed recently. The use of such a theoretical framework may enable the development of computational approaches of analysis based on results of graph theory. Here we propose a novel method to integrate miRNA and mRNA data into a single comprehensive model able to extract meaningful information. Starting from miRNA information stored in databases and from expression data of both mRNA and miRNA we build a weighted bipartite graph, called association graph representing the relations among those molecules. Then we use this graph in a two fold way. Initially, we extract maximally weighted cliques that represent association among miRNA and mRNAs. In parallel the comparison among association graphs derived from different experimental conditions, i.e. different diseases or different species, may reveal conserved mechanism and disrupted ones in different situations.