Home // BIOTECHNO 2017, The Ninth International Conference on Bioinformatics, Biocomputational Systems and Biotechnologies // View article

Cancer Gene Analysis Using Small Matryoshka (SM) Found by Matryoshka Feature Selection Method

Authors:
Shuichi Shinmura

Keywords: Cancer Gene Analysis; Cancer Gene Diagnosis; Microarray Dataset; Matryoshka Feature Selection Method; Small Matryoshka (SM); Basic Gene Set (BGS); NP-hard; large p small n.

Abstract:
We established a new theory of discriminant analysis after R. Fisher. We developed two methods and four Optimal Linear Discriminant Functions (OLDFs) in order to solve five serious problems of discriminant analysis. Over than 30 years, many researchers could not select cancer genes from gene datasets such as microarray datasets (Problem 5). The Matryoshka feature selection method (Method 2) could separate the dataset to several linearly separable subspaces (Small Matryoshka, SM) and non-linearly separable subspace definitely. We consider genes including each SM are cancer genes because they can discriminate cancer patients versus normal patients completely. On the other hand, other genes cannot discriminate two classes correctly and are noise. Therefore, Method 2 can separate the dataset to several signal SMs and noise subspace naturally. In this study, we introduce how to analyze all SMs of six datasets using common statistical methods and propose malignancy indexes for cancer gene diagnosis. Because our standard statistical approach obtains almost the same successful results, we explain the results by Alon et al. dataset. Researchers can analyze their dataset by our approach.

Pages: 1 to 9

Copyright: Copyright (c) IARIA, 2017

Publication date: May 21, 2017

Published in: conference

ISSN: 2308-4383

ISBN: 978-1-61208-560-9

Location: Barcelona, Spain

Dates: from May 21, 2017 to May 25, 2017