 The paper presents a two-stage multi-filter hybrid model of feature selection for colon cancer classification using information gain and genetic algorithm, followed by minimum redundancy maximum relevance, MRMR, technique to filter and rank genes, with promising results from decision tree, k-nearest neighbor, and night-based classifiers, achieving a higher accuracy compared to existing methods for colon cancer diagnosis and drug discovery. This article was authored by Murad Al-Rajib, Murad Al-Rajib, Joan Lu, and others.