Classification of Genes for Disease Identification Using Data Mining Techniques

Scientists are nowadays providing great awareness about microarray gene expression dataset. Researches tell that data mining fails to recognize the most important biological associations between genes. Recently biological information mining using clustering techniques were used for the analytical evaluation of gene expression. There are many challenges exist in the existing methods. Optimization Algorithm for multi dimensional search space does not provide relational optimization result on varying gene expressional problems. The existing method solves the clustering problem, but bi-cluster based gene expression information was not extracted. A key point on existing work was to handle multi modal structure optimization problems with effective searching process, but it does not offer relational sequence optimized result on the associated gene data. To overcome the above issues, the proposed research is developed. The objective of the proposed research is to extract the biological information and identify the relational sequences on gene expression to identify abnormal genes. The proposed techniques like biological process on physiological data, PCPHC and Biclustered Ant Optimized Feature Relational Sequencing extract the biological process information from gene expression datasets. These techniques are tested on various bench marked datasets called Cancer Gene Expression datasets Broad Institute repository for experimental evaluation of the proposed method with an existing method, which identifies and extracts the hidden information from datasets. Finally, the gene patterns were verified as normal or abnormal on the basis of simple pattern matching process.