A new hierarchical method for identification of dynamic regulatory pathways from time-series DNA microarray data

A new hierarchical method is proposed to analyze timeseries DNA microarray data to identify dynamic genetic pathways. Initially the hierarchical method applies a specialized clustering technique to incorporate the available heuristic information about biological system. Then, the prototypes of the resulting clusters are used as time-variables to develop an auto regressive model to relate the expression of the prototypes to each other. The resulting model also allows the prediction of gene expressions for the next time steps. The developed AR model can then be used to relate the expression value of each single gene to the genes of other clusters. The proposed method was applied to the cell-cycle dataset containing the DNA microarray time-series of a large number of genes involved in the eukaryotic cell-cycle. The technique resulted to a network of interactions among five clusters of genes in which the genes of each cluster have a biologically-meaningful trend in time.