BRCA gene expression level analysis for identification of breast cancer using computer assisted algorithms

Abstract

This paper deals with the study and analysis of gene expression data in identifying breast cancer, a second leading type of cancer in the world. Breast tumourigenesis consists of genetic changes which results in the altered mRNA and protein levels, such as the intensification of oncogenes or loss of tumor suppressor genes. In this paper, we have used a novel approach for finding the disease – causing gene using computer-assisted algorithms. The existing algorithms are compared with each other to determine the efficiency in detecting the diseases from gene expression value. The results proved that the effectiveness of Orthogonal Nonnegative Matrix Tri-Factorization (ONMTF) algorithm performs better than sequential and Divide and Conquer Kernel Solving Support Vector Machines (DCKSVM) algorithm and Hybrid Radial Bias Neural Network (HRBFNN) algorithm in finding the diseased gene.