I am interested in methods and applications of Bayesian inference. More specifically, I am working on nonparametric Bayesian inference, decision problems, and applications to biomedical research problems. Nonparametric Bayesian inference refers to prior models for infinite dimensional random quantities, typically random probability measures. Decision problems include particular clinical trial design and multiple comparison procedures. Other applications that I am interested in include inference related to dependence structure, specifically graphical models to formalize inference about dependence for high throughput genomic data. Another large area of application is population pharmacokinetic and pharmacodynamic models, which give rise to many good applications that exploit many of my methodological interests.

My undergraduate education is from Universität Wien and Technische
Universität Wien, Austria. My Ph.D. is from Purdue University where
I worked under Jim Berger on MCMC for constrained parameter
problems. I spent several years at the Institute of Statistics and Dec
Sciences (ISDS), Duke University, and at M.D. Anderson Biostatistics.

CV

Grants

Projects

CV
Academic Positions:
Professor of Statistics (UT Austin, 2011–present)
Professor of Biostatistics (M.D. Anderson Cancer Ctr. 2001–2011)
Associate and Assistant Professor of Statistics (Duke U., 1991–2002)

Honors:
Fellow of the American Statistical Association
President of the International Society for Bayesian Analysis (2010)
Robert R. Herring Distinguished Professorship in Clinical Research (2007–2011)

Research Interests:
Bayesian analysis and decision making: Markov chain Monte Carlo methods simulation based optimal design, dynamic models, networks, clinical trial design, dependent gene expression, longitudinal data models, pharmacokinetic/pharmacodynamic.

Nonparametric Bayes: semiparametric mixture models, mixture of Dirichlet process models, random partitions, clustering.

Recent Publications
(total 87 peer reviewed papers, 25 proceedings and other writings, 9 discussions, 4 edited books, 1 book)
1. Trippa, L, Rosner, G., and Müller, P. (2011), “Bayesian Enrichment Strategies for Randomized Discontinuation Trials.” Biometrics, to appear.
2. Trippa, L, Müller, P. and Johnson, W. (2011), “The Multivariate Beta Process and an Extension of the Polya Tree Model.” Biometrika, to appear.
3. Müller, P., Quintana, F, and Rosner, G. (2011), “Bayesian Clustering with Regression.” Journal of Computational and Graphical Statistics, to appear.
4. Sivaganesan, S., Laud, P. and Müller, P. (2011), “A Bayesian Subgroup Analysis with a Zero-Enriched Polya Urn Scheme.” Statistics in Medicine, to appear, DOI: 10.1002/sim.4108.
5. Jara, A., Hanson, T., Quintana, F., Müller, P., and Rosner, G. (2011) “DPpackage: Bayesian Non- and Semi-parametric Modelling in R.” Journal of Statistical Software, to appear.
6. Yang, Y., Müller, P. and Rosner, G. (2010) “Semiparametric Bayesian Inference for Repeated Fractional Measurement Data.” Chilean Journal of Statistics, 1, to appear.
7. Harvey, C., Liechty, J., Liechty, M., and Müller, P. (2010), “Portfolio Selection with Higher Moments.” Quantitative Finance, to appear.
8. Li, Y., Müller, P. and Lin, X. (2010), “Center-Adjusted Inference for a Nonparametric Bayesian Random Effect Distribution.” Statistica Sinica, to appear.
9. Morita, S., Thall, P. and Müller, P. (2010), "Evaluating the Impact of Prior Assumptions in Bayesian Biostatistics." Statistics in Biosciences, to appear.
10. Chen, Y.,A., Almeida, J.S., Richards, A.J., Müller, P., Carroll, R.J., and Roherer, B. (2010), “A nonparametric approach to detect local correlation in gene expression.” Journal of Graphical and Computational Statistics, to appear.
11. Leon-Novelo, L.G, Zhou, X, Bekele, B., and Müller, P. (2010), “Assessing Toxicities in a Clinical Trial: Bayesian Inference for Ordinal Data Nested within Categories.” Biometrics, 66, 966-74.
12. Müller, P. and Quintana, F. (2010), “Random Partition Models with Regression on Covariates.” Journal of Statistical Inference and Planning, doi:10.1016/j.jspi.2010.03.002
13. Berry, S., Carlin, B., Lee, J. and Müller, P. (2010) Bayesian Adaptive Methods for Clinical Trials, Chapman & Hall.
14. Chen, M.-H., Dey, D.K., Müller, P, Sun, D. and Ye, K. (eds.) (2010), Frontiers of Statistical Decision Making and Bayesian Analysis, Springer-Verlag, New York.
15. Hjort, N., Holmes, C., Müller, P. and Walker, S. (eds.), (2010) Bayesian Nonparametrics, Cambridge University Press.

Full CV

Grants
Current
5 R01 CA075981-12 (Mueller)
09/25/2007 – 08/31/2012
2.0 calendar
NIH/NCI
$231,689
“Population Pharmacokinetics/Dynamics: Statistical Issues”
The major goals of this project are to analyze the pharmacokinetics of two or more drugs given simultaneously, develop statistical tools to carry out full Bayesian meta-analysis across studies with correlated outcome data and optimize design rules for determining sample criteria in the context of modeling complex biologic processes.

5 R01 CA132897-03 (Ji)
09/15/2008 – 07/31/2013
2.4 calendar
NIH/NCI   
$23,160
“Bayesian models for cancer prognosis by integrating diverse types of data”
The long-range goal of this application is to improve risk predication, treatment selection, and subtype classification in cancer prevention, diagnosis, and prognosis. The short-term objective is to improve prediction of treatment response for cancer patients by developing innovative statistical models that integrate three different types of data, including two subtypes of informatics data, namely protein pathway data and high- throughput protein expression data, and a third type, which is the standard clinical and demographic data.

Pending
RC1 CA144738 (Carlin)
02/01/2011- 01/31/2014
1.0 calendar
NIH/NCI
$481,826 (subcontract)
“Statistical Methods and Software for More Efficient, Ethical, and Affordable Clinical Trials”
We develop methods to borrow strength and combine inference across related subpopulations. Specifically, we develop methods to exploit available historical data, as well as more general approaches for inference across related subpopulations. The common theme is inference across subpopulations when the subpopulations are a priori not exchangeable. There is no widely used mechanism in the literature to combine inference across non- exchangeable subgroups. This project will fill this gap in the literature by developing an approach based on random clustering of subpopulations. Work includes the development of publically distributed software.

Projects
Population PK/PD: statistical issues. We investigate the use of highly structured Bayesian models for problems arising in population PK/PD. This work is jointly with Gary Rosner, JHU.

Bayesian methods for cancer prognosis by integrating diverse types of data. We develop models and methods for Bayesian inference for a wide variety of experimental platforms. Examples include histone modifications, RPPA protein activation data and more. We are particularly interested in learning about dependence patterns, borrowing strength across platforms and learning about changes of dependence patterns. This work is jointly with Yuan Ji, UT MDACC.

Subgroup analysis: When a clinical study cannot find the desired conclusion for the original overall patient population, it is often of interest to consider inference for smaller subpopulations. Subgroups are characterized by available baseline covariates. We are exploring the use of decision theoretic and model-based Bayesian methods to approach this problem. This is joint work with Prakash Laud, Medical College Wisconsin, and Siva Sivaganesan, U. Cincinnati.

Nonparametric Bayesian inference: I am working on novel models and methods in nonparametric Bayesian inference. Several related projects are jointly with Fernando Quintana, PUCC, Santiago.