Recent advances in high-throughput experimental methods in molecular biology hold great promise. DNA microarray technologies enable researchers to measure the expression levels of thousands of genes simultaneously, and more recently microarrays have been exploited to measure genome-wide protein-DNA binding events. Time series expression data offer particularly rich opportunities for understanding the dynamics of biological processes. However, DNA expression data is noisy and often many data points are missing. Time series expression data add additional complications, including sampling rate differences between experiments and variations in the timing of biological processes. Thus, principled computational methods are required in order to make full use of time series expression data.

In this talk, I will present algorithms for analyzing time series expression data at two different levels: individual genes and genetic regulatory networks. For the first level, I will present algorithms that permit the principled estimation of unobserved time-points,clustering and the identification of differentially expressed genes. By applying these algorithms, we provide new insights into the role of two key cell transcriptional factors. For the network level, I will describe a new algorithm that efficiently combines complementary large-scale expression and protein-DNA binding data to discover co-regulated modules of genes. The discovered modules are used to build a regulatory network of transcription factors and modules, to label transcription factors as activators or repressors and to identify patterns of combinatorial regulation. Finally, I will present an algorithm which combines the above methods to automatically infer dynamic sub-networks for specific biological processes.

This talk is designed to be accessible to computer scientists. No prior biological knowledge will be assumed.