Allen D. Roses, MD

Professor of Neurology
Department / Division:
Neurology / Behavioral Neurology
Bryan Alzheimer's Disease Research Ctr
2200 West Main Street, Ste A-200
Durham, NC 27705
Office Telephone:
(919) 660-8065
  • MD, University of Pittsburgh, 1967
  • Chief Resident in Neurology, Medicine, Duke University
  • Neurology Resident, Medicine, Columbia University
Research Interests:
Allen D. Roses has established an international reputation for his work in pharmacogenetics, exploratory drug discovery, and clinical neuroscience.

Dr. Roses founded Cabernet Pharmaceuticals in 2008 to provide pharmacogenetics (PGx) and project-management services to pharmaceutical and biotechnology companies, clinical-research and managed-healthcare organizations, and academic institutions. He has formed a team of consultants with deep experience in the practical application of PGx to drug development.

Dr. Roses also serves in several capacities at Duke University: as Jefferson-Pilot Professor of Neurobiology, and Genetics, as Professor of Medicine (Neurology) as Director of the Deane Drug Discovery Institute, and as Senior Scholar at the Fuqua School of Business.

He recently returned to Duke after a decade-long career as a Senior Vice President at GlaxoSmithKline (GSK) and its corporate predecessor GlaxoWellcome (GW). Upon joining GW in 1997, he organized genetic strategies for susceptibility-gene discovery, pharmacogenetics strategy and implementation, and integration of genetics into medicine discovery and development. Subsequently at GSK, he headed research in genetics, genomics, proteomics, and bioinformatics in support of the entire R&D pipeline.

Among the specific activities of his group at GSK were proof-of-principle experiments using linkage-disequilibrium mapping to identify susceptibility loci for drug-associated adverse events. With respect to hypersensitivity to the HIV/AIDS drug abacavir, for example, GSK identified the HLA-*B5701 locus with candidate-gene analyses and then prospectively established the sensitivity (97%) and specificity (>99%) of this genetic risk marker.

During his previous tenure at Duke, Dr. Roses was Jefferson Pilot Professor of Neurobiology and Neurology, Founding Director of the Joseph and Kathleen Bryan Alzheimer’s Disease Research Center, Chief of the Division of Neurology, and Director of the Center for Human Genetics.

He was one of the first clinical neurologists to apply molecular genetic strategies to neurological diseases. His laboratory reported the chromosomal location for more than 15 diseases, including several muscular dystrophies and Lou Gehrig’s disease. He led the team that in 1992 identified APOE as a major, widely confirmed susceptibility gene in common late-onset Alzheimer’s disease. Translation of these findings to metabolic-pathway analyses and drug discovery and development continued in GSK, leading to Phase III trials now under way to evaluate the drug rosiglitazone for the treatment of Alzheimer’s disease.

Dr. Roses was a member of the Science Board of the US Food and Drug Administration between 2003-2007. He was a member of the Board’s Subcommittee on Science and Technology that in 2007 authored the report “FDA Science and Mission at Risk.” He continues to consult with the FDA and other regulatory agencies in the field of pharmacogenetics and companion diagnostics.
Recently Dr. Roses described the association of a variable polyT repeat [rs10524523] with the age of onset distribution of AD. The data enhanced the accuracy of the prior age of onset curves based solely on APOE genotypes [also developed by Dr. Roses at Duke in 1992]. Shiraz Pharma was founded in 2009 to commercialize the IP from this new discovery. Zinfandel Pharma was also founded in 2009 to plan and execute a prospective validation of the “523” diagnostic for predicting risk of AD onset in the next 5-7 years for individuals aged 60-87 years. A combination diagnostic validation study and prevention [delay of age of onset] clinical trial in epidemiologic-selected populations is currently in progress, having completed a Voluntary Genomic Data Submission discussion with the FDA regarding the design of the clinical trial. Five epidemiologic-based recruitment
end fo content area