Medulloblastoma is a common and disabling tumor of children, for which current therapies are often ineffective. The transcription factors MYC and MYCN are expressed in the majority of these tumors, but are essentially untargetable. The long-term objective is to integrate genomic screening of human tumors with mouse co-expression networks to identify both new cancer targets, and novel roles for pathways with existing therapeutics. We focus this effort in-part on tissue-specific networks that distinguish MYC and MYCN in in normal murine cerebellum, and in mouse models of MYC and MYCN-driven medulloblastoma. We will filter potential targets through analysis of murine MYC and MYCN driven networks in skin and in skin cancer, and using human genetics and genomic datasets. We present preliminary data utilizing this methodology to identify GABA receptors as novel targets in medulloblasotma. This study establishes a general method of developing murine genetic networks in normal murine somatic and cancer tissues, and applying these as a filter to human cancer, to identify new targets. Our specific aims are: A1. Develop an integrated bioinformatics screening and mouse models approach to identify cancer targets. We characterized co-expression networks in normal skin, brain, and lung from a large genotyped mouse back- cross, from which 22 tissues are preserved for trans-CTDD projects. A2. Based on data resulting from our integrated mouse-human bioinformatics approach, we will validate GABA receptors as therapeutic targets in MB, and whether bioactive drugs used in neurological and psychiatric disorders can be leveraged therapeutically in this common and lethal childhood cancer. A3. To identify targets associated with MYC and/or MYCN in medulloblastoma. Successful completion identifies MYC and MYCN-associated signature and targets in medulloblastoma, potentially applicable across a spectrum of MYC and MYCN-driven cancers. Medulloblastoma is a common and debilitating tumor of children, for which current therapies are often ineffective. The cancer genes MYC and MYCN are associated with a majority of these tumors, but are essentially untargetable by any current drug or technology. We establishe a general method to develop and analyze genetic networks in normal mouse tissues and in mouse models for cancer (focusing here on medulloblastoma), and applying these as a filter to genetic and genomic databases from human cancer, to identify new therapeutic targets.