Summary Despite tremendous progress in clinical genomics, nearly half of children with genetic disease never receive a diagnosis. One reason is that patients are initially diagnosed based on symptoms and physical manifestations but for many genetic conditions, those symptoms can vary dramatically even for patients with the same mutation. When patients with the same mutation have different phenotypes, that also has implications for treatment and prognosis, even when a molecular diagnosis is secured. The long-term goal of our research program is to understand how genotype is translated to variation in phenotype using a variety of model systems. In our previous funding cycle, we used an allelic series of Fibroblast growth factor 8 (Fgf8) mice and showed that levels of Fgf8 gene expression and downstream signalling factors were non-linearly related to the phenotype. Little change in phenotype occurs until Fgf8 drops below 50% compared to the wildtype, and then below that level small changes in Fgf8 produce large changes in the mean phenotype and its variance. The next step in our research advances this concept and helps generate a more mechanistic understanding of variable penetrance and robustness. Specifically, we propose two conceptually related aims that address distinct concepts in understanding the central question of how mutations perturb development to generate varied phenotypic outcomes. In specific Aim 1, we test mechanisms by which gene interactions within a pathway generate variable penetrance and expressivity. Our previous research has identified epistatic interactions that may modulate FGF8, and in this Aim, we test these interactions directly. In Specific Aim 2, we test how a mutation that affects a specific but ubiquitous process generates a multi-modal range of phenotypes. This is important because genetic diseases often involve disruptions to mechanisms that operate across diverse tissues and at many points in development and yet result in highly specific birth defects such as holoprosencephaly or cleft lip and palate. As a model for this general phenomenon, we investigate the effects of a null mutation in Nitric Oxide Synthase Interacting Protein (Nosip) in mice. NOSIP is an E3-monoubiquitin ligase that modifies activity of target proteins with diverse developmental roles. In both aims, we employ a highly innovative combination of imaging, morphometric, transcriptomic and experimental approaches to integrate multiple levels of input (genetic, molecular, cellular) to understand how the effects of a mutation cascade across levels from gene expression to cellular dynamics to influence organismal form.