Merge Method for Shape-Based Clustering in Time Series Microarray Analysis
A challenging task in time-course microarray data analysis is to combine the information provided by multiple time series in order to cluster genes meaningfully. This paper proposes a novel merge method to accomplish this goal obtaining clusters with highly correlated genes. The main idea of the proposed method is to generate a clustering, starting from clusterings created from different time series individually, that takes into account the number of times each clustering assemble two genes into the same group. Computational experiments are performed for real-world time series microarray with the purpose of finding co-expressed genes related to the production and growth of a certain bacteria. The results obtained by the introduced merge method are compared with clusterings generated by time series individually and averaged as well as interpreted biologically.
Keywordsmicroarray analysis time series clustering merge methods
Unable to display preview. Download preview PDF.
- 4.Wei, J.S., Greer, B.T., Westermann, F., Steinberg, S.M., Son, C.-G., Chen, Q.-R., Whiteford, C.C., Bilke, S., Krasnoselsky, A.L., Cenacchi, N., Catchpoole, D., Berthold, F., Schwab, M., Khan, J.: Prediction of clinical outcome using gene expression profiling and artificial neural networks for patients with neuroblastoma. Cancer Research 64(19), 6883–6891 (2004)CrossRefGoogle Scholar
- 5.Coffey, N., Hinde, J.: Analyzing time-course microarray data using functional data analysis - a review. Statistical Applications in Genetics and Molecular Biology 10 (2011); peer-reviewedGoogle Scholar
- 8.Storey, J., Xiao, W., Leek, J., Tompkins, R., Davis, R.: Significance analysis of time course microarray experiments. UW Biostatistics Working Paper Series (August 2004)Google Scholar