Finding synergy networks from gene expression data: A fuzzy-rule-based approach

Article Type

Research Article

Publication Title

IEEE Transactions on Fuzzy Systems

Abstract

Genes interact among themselves directly as well as indirectly, and thereby, a gene regulates the expression levels of other genes. In this work, our objective is to identify a special type of network called 'synergy network.' We want to find synergistic gene pairs that interact via collaboration with respect to a disease and form a network of such synergistic genes. First, we discuss some issues related to existing information-theoretic methods of finding synergy networks and, then, propose a fuzzy-rule-based approach for discovery of synergy networks. We justify that fuzzy rule base is a natural choice to realize all the desired attributes of synergistic relations. To our knowledge, this is the first attempt to exploit fuzzy modeling for finding synergy networks. The system uses a set of human understandable rules that is generated at a low cost for every pair of genes. We apply our method on two prostate cancer datasets. We show that the proposed method is capable of discovering gene pairs that collaborate with each other with respect to prostate cancer. We demonstrate that our results are statistically significant. We also discuss the relevance of the identified genes to cancer biology.

First Page

1488

Last Page

1499

DOI

10.1109/TFUZZ.2016.2540062

Publication Date

12-1-2016

Link to Full Text

Share

COinS