Extraction of Co-Expressed Degs From Parkinson Disease Microarray Dataset Using Partition Based Clustering Techniques

Data mining has been done a great progress in the recent years. It is helpful to acquiring knowledge from large domains of databases. In data mining clustering is extremely useful to find distribution design in the fundamental data. Clustering algorithm generallyemploy as distance metric constructedsimilarity measure in order to cluster the databases such that data points in the same partition are more similar than points in different partition. The proposed research work focuses on Computing DEGs, Up and Down Regulated DEGs from Parkinson Disease Microarray Gene expression data set using Statistical significance and Biological significance, Classifying the Up and Down Regulated DEGs using the K-means, PAM and CLARA partitioning clustering algorithm, and validating the cluster results,from the results it is observed that K-means performs well for microarray datasets. Clustering itself is not enough to identify the significant similarity of genes biomarkers with respect to the particular disease category. Further it needs to identifying the Co-expressed gene network to get deep insights to extract significant DEGs which really part of the particular disease. Thus, Co-expressed genes are identified using Pearson Correlation method. From the proposed research work it is observed that DEGs clustered by K-means algorithm have more interaction with the Co-experession raio 0.39 have more interaction, CLARA algorithm with the Co-experession raio 0.38 and the PAM algorithm with the Co-experession raio 0.32 have the less interaction.