Fuzzy Rule Based Clustering for Gene Expression Data

The complexity of biological networks and the large number of genes in microarray datasets cause a lot of challenges in analyzing gene expression data. Clustering techniques which group the similar genes into the same clusters with the purpose of analyzing the function of genes have been used to overcome these challenges. In general, fuzzy clustering methods are more suitable for analyzing gene expression data because of overlap between the biological groups and existing noisy data within the microarray datasets. In this paper by the usage of FRBC(Fuzzy Rule Based Clustering algorithm) approach a fuzzy clustering algorithm is proposed to automatically explore the potential gene clusters in the microarray datasets with no prior knowledge and represent them with some interpretable fuzzy rules that are human understandable. In the simulation results, the accuracy of the algorithm is evaluated on some microarray datasets and to confirm whether the clusters are precisely explored, several validity criteria are used to compare the proposed algorithm with some well-known fuzzy clustering methods.