Knowledge Discovery in the Identification of Differentially Expressed Genes in Tumoricidal Macrophage
High-throughput microarray data are extensively produced to study the effects of different treatments on cells and their behaviours. Understanding this data and identifying patterns of groups of genes that behave differently or similarly under a set of experimental conditions is a major challenge. This has motivated researchers to consider multiple methods to identify patterns in the data and study the behaviour of hundreds of genes. This paper introduces three methods, one of which is a new technique and two are from the literature. The three methods are cluster mapping, Rank Products and SAM. Using real data from a number of microarray experiments comparing the effects of two very different products that can activate macrophage tumoricidal activity we have identified groups of genes that share interesting expression patterns. These methods have helped us to gain an insight into the biological problem under study.
Unable to display preview. Download preview PDF.
- 1.Brazma, A., Vilo, J.: Gene expression data analysis. Federation of European BiochemicalSociety 480, 17–24 (2000)Google Scholar
- 5.Famili, A., Liu, Z., Ouyang, J., Walker, R., Smith, B., O’Connor, M., Lenferink, A.: A novel data mining technique for gene identification in time-series gene expression data. In: ECAI Workshop on Data Mining in Genomics and Proteomics, pp. 25–34 (2003)Google Scholar
- 8.Tang, C., Li, Z., Zhang, A., Ramanathan, M.: Interrelated two-way clustering: an unsupervised approach for gene expression data analysis. In: Proceedings of the 2nd International Symposium on Bioinformatics and Biocomputing, pp. 41–48 (2001)Google Scholar
- 12.Wolpert, D.H.: Stacked Generalization. In: Neural Networks, vol. 5, pp. 241–259. Pergamon Press, Oxford (1992)Google Scholar