Biotechnology Letters

, Volume 27, Issue 8, pp 597–603

A gene selection algorithm based on the gene regulation probability using maximal likelihood estimation

Article

DOI: 10.1007/s10529-005-3253-0

Cite this article as:
Wang, HQ. & Huang, DS. Biotechnol Lett (2005) 27: 597. doi:10.1007/s10529-005-3253-0

Abstract

A novel gene selection algorithm based on the gene regulation probability is proposed. In this algorithm, a probabilistic model is established to estimate gene regulation probabilities using the maximum likelihood estimation method and then these probabilities are used to select key genes related by class distinction. The application on the leukemia data-set suggests that the defined gene regulation probability can identify the key genes to the acute lymphoblastic leukemia (ALL)/acute myeloid leukemia (AML) class distinction and the result of our proposed algorithm is competitive to those of the previous algorithms.

Keywords

DNA microarray data-set gene expression gene regulation probability gene selection maximum likelihood estimation 

Copyright information

© Springer 2005

Authors and Affiliations

  1. 1.Intelligent Computation Lab, Hefei Institute of Intelligent MachinesChinese Academy of ScienceHefei,China
  2. 2.Department of AutomationUniversity of Science and Technology of ChinaHefeiChina