Gene expression based Computer Aided diagnostic system for Breast Cancer: A novel biological filter for biomarker detection

Abstract

Cancer is a complex disease because it makes complex cellular changes. Therefore, microarrays have become a powerful way to analyse cancer and identify what changes are produced within a cell. Through DNA microarrays, it has become possible to look at the expression of thousands of genes in one sample and this is called gene expression profiling. Gene expression profiling is important to capture a set of expressed genes that determines a cell phenotype. However, analysing microarray data is challenged by the high-dimensionality of the data compared with the number of samples. The aim of this study was to enhance the diagnostic accuracy of Breast Cancer Computer Aided Diagnostic Systems (CADs) that use gene expression profiling of peripheral blood cells, by introducing a novel feature selection method called Bi-biological filter that was further refined by Best First Search with Support Vector Machines SVM (BFS-SVM) to select a small set of the most effective genes predictive of breast cancer. From each patient's gene expression profiles, a gene co-expression network was built and divided into functional groups or clusters using Topological Overlap Matrix (TOM) and Spectral Clustering (SC) in the design of the Bi-Biological filter to obtain the preliminary set of gene markers. BFS-SVM was used to further filter a smaller set of best gene markers, and Artificial Neural Networks (ANN), SVM and Linear Discriminant Analysis (LDA) were used to assess their classification performance. The study used 121 samples - 67 malignant and 54 benign cases as input to for the system. The Bi-biological filter selected 415 genes as mRNA biomarkers and BFS-SVM was able to select just 13 out of 415 genes for classification of breast cancer. ANN was found to be the superior classifier with 93.4% classification accuracy which was a 14% improvement over the past best CAD system developed by Aaroe et al. (2010).