Project Summary Our overall goal is to improve outcomes for children with forms of cancer that cannot be eradicated with current therapies. We are focused on neuroblastoma, an especially challenging form of childhood cancer that accounts for a large proportion of childhood cancer deaths each year, and we have assembled a team to explore a new approach in which a novel computational pipeline is applied to existing genomics data and a functional genomics screen to reveal new insights into neuroblastoma biology. We anticipate that new biomarkers for risk stratification and assignment of molecularly targeted therapy will stem from our work. In the US each year, over 700 children and young adults develop neuroblastoma, among the most common solid malignancies in children. Sadly, the chance of cure is low for those with high-risk disease, and this bleak outlook has only modestly improved with the application of multifaceted therapies in recent years. New molecular biology and molecular genetics tools at the close of the last century brought new insights into the underpinnings of neuroblastoma, including the fact that copy-number gain in the MYCN gene is among the most important determinants of biologic risk. However, that has not been translated into better treatment and other molecular derangements contribute to poor chances of survival for children with this disease. Many clinicians, scientists, and patients and their parents anticipated that the more recent genomics revolution would usher in “precision” medicine focused on the mutant forms of proteins anticipated to drive the disease. That promise has not been fully realized in cancers like neuroblastoma that lack highly-recurrent, targetable mutations. Our team came together to explore a new approach to help close this gap. Given the few recurrent mutations in this disease, we are considering neuroblastoma to be a cancer in which normal developmental programs are corrupted by altered gene expression and that the altered gene expression is often “hard-wired” into the cell by gains and losses in the copies of the genes encoding oncogenic drivers and tumor suppressors. We exploring the capacity for a new computational algorithm to identify those cancer drivers/suppressors using existing genomic datasets. Second, we propose to use a focused but high-throughput cell-based screen to quickly provide functional validation of the candidate neuroblastoma drivers. Finally, we are using this information to develop a new biologically-based tool for assigning risk and guiding treatment assignment for children with neuroblastoma. If successful, we can extend this developmental model to other forms of childhood cancer.