The significance and impact of melanoma as a disease entity can not be understated. Despite the long history of clinical and molecular efforts directed towards this disease, surprisingly little is known about the precise genetic lesions leading to melanoma and even less is known with regard to how these few genetic lesions relate to disease classification or progression. Significant progress on both the basic and clinical fronts could could be achieved through the production of an accurate mouse model of malignant melanoma that faithfully reproduces disease progression on the pathological and molecular levels. This proposal attempts to refine and validate further an established mouse model of cutaneous melanoma. To achieve this goal, mice will be engineered to possess several genetic lesions commonly observed in human melanomas, including activated MET, EGF receptors as well as disruption of the p16ink42, PTEN and possibly Mxil genes. Evolving gene expression patterns and genomic changes at various tumor stages will be extensively cataloged as a means of validation. This refined model of melanoma should serve to advance our understanding of melanoma biology as well as to provide a system for melanoma gene discovery. The latter will include a combination of CGH, genome wide LOH, genetic mapping of susceptibility loci and candidate gene mutational analyses. The use of these melanoma mice in preclinical testing are outlined as well.