Som-Based Class Discovery Exploring the ICA-Reduced Features of Microarray Expression Profiles
2004
Dragomir, Andrei(Medical School, University of Patras) | Mavroudi, Seferina(Medical School, University of Patras) | Bezerianos, Anastasios(Medical School, University of Patras)
Gene expression datasets are large and complex, having many variables and unknowninternal structure. We apply independent component analysis (ICA) to derive aless redundant representation of the expression data. The decomposition producescomponents with minimal statistical dependence and reveals biologically relevantinformation. Consequently, to the transformed data, we apply cluster analysis (animportant and popular analysis tool for obtaining an initial understanding of thedata, usually employed for class discovery). The proposed self-organizing map(SOM)-based clustering algorithm automatically determines the number of ‘natural’subgroups of the data, being aided at this task by the available prior knowledge of thefunctional categories of genes. An entropy criterion allows each gene to be assignedto multiple classes, which is closer to the biological representation. These features,however, are not achieved at the cost of the simplicity of the algorithm, since themap grows on a simple grid structure and the learning algorithm remains equal toKohonen’s one.
Показать больше [+] Меньше [-]Библиографическая информация
Эту запись предоставил Hindawi
