Bicluster method to predict gene patterns to classify differential gene expressions in non-small cell lung cancer

Abstract

In recent years, there are numerous efforts to overcome the constraints of data mining approaches to classify "BIG DATA". There are several types of data which has identical and similar expressions but there are dependent classification algorithms to predict these classes of expressions. Totally Different algorithms have been developed and enforced to research and differentiate the categories of data groups based on their functions. Zero-suppressed Binary Decision Diagram (ZBDD) algorithms help to classify the data with several categories. In the present study, lung cancer gene expression datasets 25 samples contain 10 mouth buccal cavity epithelial tissue samples and 15 nasal epithelial tissue samples from never smokers and current smokers were used to classify the genes and their expressions with various conditions. Using R and BioConductor software to normalize and predict differential expressed genes by Affy, Affycore tools and Limma packages to predict the gene expression with various functional properties. ZBDD algorithm and parallel coordination helps to predict the functional genes and the results shows 345 nasal epithelial genes were predict of which 54 genes were present in bicluster and 35 genes from mouth epithelial tissues show 14 were present in ZBDD bicluster. The results conclude that ZBDD algorithm has great advantage to classify big data and this algorithm can introduce in any large datasets for the accurate predict of large datasets.