Practical Screening Method for Cancer Gene Diagnosis, How to Choose Cancer and Normal Patients by Four Principles

Abstract

We developed a new theory of discriminant analysis (Theory1). Physicians can use it for practical medical� diagnoses. Only Revised IP Optimal-LDF (RIP) obtains the minimum number of misclassification (MNM). RIP can� discriminate linearly separable data (LSD) theoretically. It discriminated against 169 microarrays with two classes and found� that 169 MNMs are zero and LSD. It can split high-dimensional arrays into many small LSD with less than n (patient�s� number) genes that are the candidates of multivariate oncogenes. We completed a new theory of high-dimensional gene data� analysis (Theory2). A 100-fold Cross-Validation (Method1) can rank all candidates for the importance of diagnosis. Thus, if� physicians firstly use Theory2 as the screening method, they can start their medical studies with the correct small sizes of� candidates. This paper analyzes four arrays in detail and proposes correctly choosing cancer and normal patients using four� principles.�