PROJECT ABSTRACT Cells choose transcriptional programs, in part, by passing information from molecular sensors to signaling cascades that modulate the activity of DNA-binding transcription factors (TFs). Changes in TF activity lead to changes in the transcription rates of specific genes, which are often the first steps in responding to new information. The molecular machinery responsible for this can be thought of as the cell's control circuits. My research program focuses on developing computational and molecular methods that make it possible to map out a cell's control circuits, to watch them as they respond to new information, and ultimately to rewire them in ways that contribute to human health and well-being. Saccharomyces cere- visiae (yeast) is the ideal organism for developing these methods because of its relatively simple ge- nome, extensive collections of strains engineered for experimental systems biology, and comprehensive datasets for testing and optimizing new methods. The research proposed here focuses exclusively on yeast, but our methods will be immediately applicable to fungal pathogens and ultimately adaptable for model organisms and humans. The first objective of our plan is to optimize methods for determining which genes are regulated by each TF. We will use these methods to produce a map of the TF-target network that goes significantly beyond what is known today, both in accuracy and completeness. Like the map of metabolic reactions, this will be a valuable resource for addressing many scientific questions. Our second objective is to develope methods for inferring the activity levels of all TFs in any sample of cells by analyzing their transcriptomes. One product of this work will be easy-to-use software that will enable other scientists to identify changes in TF activity in any set of yeast transcriptional profiles. Our third objective is to develop methods for identifying proteins that regulate the activities of each TF. This will make it possible to explain the changes in TF activity we observe when stimuli, such as drugs or nutrients, are provided to cells, and to design experiments that test those explanations. Achieving these objectives will make it possible to approach our ultimate goal – to develop a quan- titative model that can predict the transcriptional response to genetic and environmental perturbations. As a concrete benchmark for success, this model should accurately predict the effect on the entire yeast transcriptome when combinations of TFs are simultaneously perturbed (deleted or overex- pressed) under growth conditions for which we have no perturbation data. We can achieve this ambi- tious, long term goal only with stable MIRA funding.