Integrating a fuzzy k-means clustering and a bayesian network approaches for identifying gene-gene interaction

Abstract EN

DNA Microarray provides a powerful basis for analysis of gene expression.

Data mining methods, such as clustering have been widely applied to microarray data to link genes that show similar expression patterns.

However, this approach usually fails to unveil gene-gene interactions in the same cluster.

In the paper, we propose (FDBN) “Fuzzy Dynamic Bayesian Network” which is a model, combining a Bayesian network with fuzzy k-means clustering in order to identify gene-gene interactions with reduced computational time compared with existing DBN methods; our approach limits potential regulators to those genes within the same cluster and with specific membership to the target gene.

To study the feasibility of our model, we worked on yeast cell cycle time-series gene expression data (116 gene) and (17 time point) with time intervals (10 min) to interpret the interactions between genes, analyze gene expression data, identify temporal events from dynamic gene expressions, and explore the reasons for diseases.