Erkan O. Buzbas
Erkan O. Buzbas
- B.S. Chemistry, Bogazici University, 2000.
- M.S. Environmental Sciences, Bogazici University, 2003.
- M.S. Statistics, University of Idaho, 2007
- Ph.D. Bioinformatics and Computational Biology, University of Idaho
- Computational Statistics
- Bayesian Statistics
- Monte Carlo Methods
- Statistical Population Genetics
After finishing my PhD, I was a postdoctoral fellow at the Department of Genetics, University of Michigan (2009-2011), and at the Department of Biology, Stanford University (2011-2012). My collaborators include my postdoctoral advisor Noah Rosenberg at Stanford University, Paul Verdu at CNRS--France, and my PhD advisor Paul Joyce at the University of Idaho.
- E.O. Buzbas. On 'Estimating species trees using approximate Bayesian computation', Molecular Phylogenetics and Evolution, 65: 1014-1016
- Paul Joyce, Alan Genz and E.O. Buzbas. Efficient simulation methods for a class of nonneutral population genetics models. Journal of Computational Biology, 19: 650-661.
- E.O. Buzbas, P. Joyce and N.A. Rosenberg. Inference on balancing selection for epistatically interacting loci. Theoretical Population Biology, 79: 102-113, Issue 3.
- Mosher, J.T., Pemberton, J.T., Harter, K., Wang, C., Buzbas, E.O., Dvorak, P., Simón, C., Morrison, S.J. and Rosenberg, N.A. “Lack of population diversity in commonly used human embryonic stem-cell lines.” New England Journal of Medicine, 362: 183-185.
My research involves theoretical modeling of evolutionary phenomena at the population level and development and applications of computational statistical methods to perform inference about evolutionary phenomena using population genetic data. I maintain a broad interest in statistical theory and philosophy of statistics, evolutionary theory, and complex systems. For more information please visit my personal website.
Currently, I focus on inference under statistical models with computationally intractable likelihoods. In particular, I work on theory and applications of approximate Bayesian computation (ABC), a class of computational statistical methods to perform inference from models with computationally intractable likelihoods. I maintain a website to keep track of developments related to the ABC methods. The website is meant to be a resource both for biologists and statisticians who want to get familiar with ABC methods and contains a short introduction to ABC, meeting announcements, and a comprehensive list of publications.