PROJECT SUMMARY/ABSTRACT The timely and precise progression of a genetic program along a cascade of regulatory steps is critical to execute a developmental process. However, gene expression is a highly stochastic process due to inevitable fluctuations in the kinetics of complex biochemical reactions; this randomness leads to substantial cell-to-cell variability (gene expression noise). The resulting phenotypic fluctuations can only be detected and quantified at the single cell level within isogenic populations. One of the most intriguing questions in science is how developmental pattern formation is executed so precisely and reproducibly despite these unavoidable fluctuations in gene expression. Presumably, mechanisms that buffer stochastic gene expression must exist. Vertebrate somitogenesis provides a paradigm system for studying this question. Somite segments (the embryonic precursors of vertebrae) are produced sequentially and periodically from the presomitic mesoderm (PSM) at the tail end of the embryo. The period of somite segmentation is controlled by the segmentation clock. The segmentation clock exhibits oscillatory expression of Hes/her-family “clock” genes due to an autoinhibitory intracellular negative feedback loop. Oscillating Delta ligands activate Notch receptors in neighboring cells and establish an intercellular positive feedback loop that synchronizes oscillation phases among neighboring cells. Disruption of these synchronized oscillations results in birth defects. The time-course of somite segmentation and epithelization occur along the posteroanterior direction in the PSM. The coordinated expression of multiple genes along the PSM are controlled by three interconnected signaling gradients (Fgf, Wnt and retinoic acid). Somitogenesis is both precise – embryos of a given species develop certain number of segments with species-specific rhythmicity – and versatile –total number of segments and their periodicity vary widely among species. Somitogenesis is also robust as embryos form segments with a certain size distribution, scaling the sizes of segments with body size, even when total cell numbers, cell sizes or growth rates are altered experimentally. These characteristics indicate that the expression noise within the oscillating segmentation network is efficiently buffered. Our overarching goal is to decipher how expression noise in gene regulatory networks is buffered during developmental pattern formation. We aspire to reach a mechanistic understanding of this buffering by combining mathematical/computational/statistical modeling with different genetic and chemical perturbations to modify dosage of multiple genes or modulate signal feedback strength.