Cytogenetic studies indicate the presence of novel amplified DNAs specific for human breast cancers. Previously,MYC, neu and PRAD1 have been found amplified in 5-40% of breast cancer biopsies and cell lines, and a number ofothers are amplified less frequently. Laborious approaches to clone DNA from the cytogenetic manifestations of gene amplification have not proven to be efficient or simple means to novel oncogenes. RDA or representational difference analysis is a newly developed molecular method that allows directed cloning of the molecular differences between two closely related populations of DNA molecules. Amplified DNA from breast and melanoma cancer cell lines has been cloned by RDA and two unique RDA probe sets for tumor analysis have been recovered. This Phase I study will explore the following issues: a) do current probes arise from novel genetic map locations and do these encode regions containing new oncogenes?; b) how can the RDA methods be optimized for cloning amplified DNA from human tumors, and c) are tumor DNAs detected by RDA probes amplified frequently in uncultured human tumor cells, and does this amplification correlate with clinical parameters? In Phase II the informative probes will be clinically cataloged and the collection of RDA probes for sites amplified in human tumors will be expanded.National Cancer Institute (NCI)