Identification of biomarker genes for cancer classification using high dimensional microarray data

Abstract

Cancer classification is an important area of research in the field of newlinebioinformatics. Researchers use microarray technology to analyze the newlineexpression levels of a large number of genes. This analysis facilitates the newlineaddressing of the issues in cancer classification and paves a way for newlinepersonalized medicine. The identification of biomarker genes pertaining to newlinecancer is a challenging task. The microarray gene expression data has a large newlinenumber of genes with different expression levels. Analyzing and classifying newlinedatasets with the entire gene space is quite difficult because there are only a newlinefew genes that are informative. The identification of biomarker genes is newlinesignificant because it improves the diagnosis of cancer disease and to make it newlinepossible to suggest personalized medicine accordingly. newlineThe main objective of the present research is to identify the newlinebiomarker genes from raw gene expression data to improve the accuracy of newlinecancer classification. Based on literature, biomarkers are identified through newlinestatistical measure or optimization algorithms. A two-stage approach is newlinefollowed to identify the biomarker genes. During the first stage, a statistical newlineapproach is applied to select the feature subset. In the second stage, an newlineoptimization algorithm is applied on the feature subset obtained from first newlinestage to identify the informative genes, which improves the classification newlineaccuracy. newlineIn the current research, the supervised machine learning techniques newlinelike Support Vector Machine (SVM) and k-Nearest Neighbour (kNN) are newlineused to construct the classification model. newline