The 16th European Conference on Artificial Intelligence (ECAI 2004), August 22-27, 2004, Valencia, Spain
data mining; genomics; gene identifications; gene expression; time-series; microarray
The purpose of this study was to develop a method for identifying useful patterns in gene expression time-series data. We have developed a novel data mining approach that identifies interesting patterns. The method consists of a combination of data pre-processing as well as unsupervised and supervised learning techniques. To evaluate our approach, we have analyzed three time series data sets which investigate the temporal transcriptome changes that occur during: 1) the cell cycle of budding yeast (<em>S. cerevisiae</em>) , 2) the epithelial to mesenchymal transition induced by Transforming Growth Factor-?1 in mouse mammary epithelial BRI-JM01 cells, and 3) the program of differentiation induced by retinoic acid in human embryonal teratocarcinoma NT-2 cells. We present the results from all of our experiments, discuss the patterns discovered through the use of our approach and briefly explain future plans and directions for improving our method.
The 16th European Conference on Artificial Intelligence (ECAI 2004) [Proceedings].