8. Selection of Discriminative Genes from Microarray Data

  1. Pradipta Maji1 and
  2. Sankar K. Pal2

Published Online: 17 FEB 2012

DOI: 10.1002/9781118119723.ch8

Rough-Fuzzy Pattern Recognition: Applications in Bioinformatics and Medical Imaging

Rough-Fuzzy Pattern Recognition: Applications in Bioinformatics and Medical Imaging

How to Cite

Maji, P. and Pal, S. K. (2012) Selection of Discriminative Genes from Microarray Data, in Rough-Fuzzy Pattern Recognition: Applications in Bioinformatics and Medical Imaging, John Wiley & Sons, Inc., Hoboken, NJ, USA. doi: 10.1002/9781118119723.ch8

Author Information

  1. 1

    Machine Intelligence Unit, Indian Statistical Institute, Kolkata, India

  2. 2

    Center for Soft Computing Research, Indian Statistical Institute, Kolkata, India

Publication History

  1. Published Online: 17 FEB 2012
  2. Published Print: 27 JAN 2012

Book Series:

  1. Wiley Series on Bioinformatics: Computational Techniques and Engineering

Book Series Editors:

  1. Yi Pan and
  2. Albert Y. Zomaya

Series Editor Information

  1. Machine Intelligence Unit, Indian Statistical Institute, Kolkata, India

ISBN Information

Print ISBN: 9781118004401

Online ISBN: 9781118119723

SEARCH

Keywords:

  • fuzzy equivalence partition matrix (FEPM);
  • gene selection;
  • information measures;
  • microarray data;
  • probability density function

Summary

This chapter establishes the effectiveness of the fuzzy equivalence partition matrix (FEPM) for the problem of gene selection from microarray data and compares its performance with some existing methods on a set of microarray gene expression data sets. It first briefly introduces various evaluation criteria used for computing both the relevance and redundancy of the genes. To measure both gene-class relevance and gene-gene redundancy using information theoretic measures such as entropy, mutual information, and f-information measures, the true density functions of continuous-valued genes have to be approximated. The chapter presents several approaches to approximate the true probability density function for continuous-valued gene expression data. It then describes the problem of gene selection from microarray data sets using information theoretic approaches. Finally, the chapter reports a few case studies and a comparison among different approximation methods.

Controlled Vocabulary Terms

fuzzy logic; numerical analysis; probability density function