Gene expression rule discovery and multi-objective ROC analysis using a neural-genetic hybrid

Microarray data allows an unprecedented view of the biochemical mechanisms contained within a cell although deriving useful information from the data is still proving to be a difficult task. In this paper, a novel method based on a multi-objective genetic algorithm is investigated that evolves a near-optimal trade-off between Artificial Neural Network (ANN) classifier accuracy (sensitivity and specificity) and size (number of genes). This hybrid method is shown to work on four well-established gene expression data sets taken from the literature. The results provide evidence for the rule discovery ability of the hybrid method and indicate that the approach can return biologically intelligible as well as plausible results and requires no pre-filtering or pre-selection of genes.

[1]  Antonio Martínez,et al.  Expression of the Interferon Regulatory Factor 8/ICSBP-1 in Human Reactive Lymphoid Tissues and B-cell Lymphomas: A Novel Germinal Center Marker , 2008, The American journal of surgical pathology.

[2]  Ed Keedwell,et al.  Genetic Algorithms for Gene Expression Analysis , 2003, EvoWorkshops.

[3]  Sayan Mukherjee,et al.  Molecular classification of multiple tumor types , 2001, ISMB.

[4]  Jonathan E. Fieldsend,et al.  Multi-class ROC analysis from a multi-objective optimisation perspective , 2006, Pattern Recognit. Lett..

[5]  N. Barnes,et al.  Serotonin: a real blast for T cells , 2007 .

[6]  B. Spiegelman,et al.  Adipsin and complement factor D activity: an immune-related defect in obesity , 1989, Science.

[7]  Jian-yong Li,et al.  Prognostic significance of ATM and TP53 deletions in Chinese patients with chronic lymphocytic leukemia. , 2008, Leukemia research.

[8]  Bþ KHI,et al.  Classification of Two-Class Cancer Data Reliably Using , .

[9]  Yusen Liu,et al.  Regulation of innate immune response by MAP kinase phosphatase-1. , 2007, Cellular signalling.

[10]  David Page.  Comparative Data Mining for Microarrays : A Case Study Based on Multiple Myeloma , 2002 .

[11]  Filip Zelezný,et al.  An experimental test of Occam’s razor in classification , 2011, Machine Learning.

[12]  R. K. Ursem.  Multi-objective Optimization using Evolutionary Algorithms , 2009 .

[13]  A. R. McNamara,et al.  A novel four zinc-finger protein targeted against p190(BcrAbl) fusion oncogene cDNA: utilisation of zinc-finger recognition codes. , 2000, Nucleic acids research.

[14]  Kalyanmoy Deb,et al.  A fast and elitist multiobjective genetic algorithm: NSGA-II , 2002, IEEE Trans. Evol. Comput..

[15]  K. Fujimoto,et al.  Glyceraldehyde-3-phosphate Dehydrogenase Enhances Transcriptional Activity of Androgen Receptor in Prostate Cancer Cells* , 2007, Journal of Biological Chemistry.

[16]  W. Lu,et al.  Regulation of hematopoietic‐specific G‐protein Gα15 and Gα16 by protein kinase C , 2003 .

[17]  K. Praveen Kumar,et al.  Memetic NSGA - a multi-objective genetic algorithm for classification of microarray data , 2007, 15th International Conference on Advanced Computing and Communications (ADCOM 2007).

[18]  T. Umemura,et al.  Cutaneous lymphoplasmacytic lymphoma with systemic metastasis in a cat. , 2011, The Journal of veterinary medical science.

[19]  Tomoyuki Hiroyasu,et al.  Discussion of Search Strategy for Multi-objective Genetic Algorithm with Consideration of Accuracy and Broadness of Pareto Optimal Solutions , 2008, SEAL.

[20]  Enrique Alba,et al.  Sensitivity and specificity based multiobjective approach for feature selection: Application to cancer diagnosis , 2009, Inf. Process. Lett..

[21]  Ash A. Alizadeh,et al.  Distinct types of diffuse large B-cell lymphoma identified by gene expression profiling , 2000, Nature.

[22]  A. Rosenwald,et al.  Disturbed expression of the T-cell receptor/CD3 complex and associated signaling molecules in CD30+ T-cell lymphoproliferations , 2010, Haematologica.

[23]  J. Mesirov,et al.  Molecular classification of cancer: class discovery and class prediction by gene expression monitoring. , 1999, Science.

[24]  M. Karin,et al.  Is NF-κB2/p100 a direct activator of programmed cell death? , 2002 .

[25]  D. Botstein,et al.  Cluster analysis and display of genome-wide expression patterns. , 1998, Proceedings of the National Academy of Sciences of the United States of America.

[26]  S. Sukumar,et al.  Role of Homeobox Genes in Normal Mammary Gland Development and Breast Tumorigenesis , 2003, Journal of Mammary Gland Biology and Neoplasia.