Engineering Gene Expression Our inability to predict the expression pattern of a gene given the sequence of its promoter is a critical problem. This inability is a direct result of our lack of understanding of the combinatorial interactions between cis-acting sequences in eukaryotic promoters. While the technological and computational tools born out of the genome project have led to good progress in identifying cis-regulatory sites, we are now shifting our focus towards understanding the interactions between those sites. Knowing which combinations of cis-regulatory sites drive which expression patterns would allow us to identify true promoters with specific activities from the thousands of "pseudo-promoters" in the genome that contain cis-regulatory sites but seem to be ignored by the cell. Many advances in biotechnology will depend on our ability to manipulate gene expression in novel ways. Decoding the rules that govern the interactions between regulatory sequences will enable us to accurately construct synthetic gene networks with useful expression properties not necessarily found in nature. This ability will empower applications based on manipulating gene expression including engineering the production of useful metabolites, driving stem cell differentiation towards specific cell fates, engineering better and more specific gene therapy vectors, and constructing accurate biosensors. Towards this end we are developing methods to quantitatively assay the expression of thousands of synthetic promoters. By applying these methods to both yeast, and mouse embryonic stem cells we deduce the additive and epistatic interactions between specific sets of cis-regulatory sequences. We use the data to construct predictive mathematical models that are then explicitly tested and refined by directed experimentation. The result of this line of investigation will be the ability to create custom promoters that drive expression patterns not found in nature, which will empower applications that depend on manipulating gene expression.