ICSA 2009 APPLIED STATISTICS SYMPOSIUM

June 21st-24th, San Francisco, California, USA

Round Table Discussion

	Topics	Leader	Date
1.	NIH Support for Statistical Research	Hulin Wu, Univ of Rochester	Monday, June 22^rd
2.	DNA copy number analysis of high throughput genomic/SNP array	Ke Zhang, Abbott	Monday, June 22^nd
3.	Statistical Opportunities in Emerging Countries	Ashwini Mathur, Novartis	Monday, June 22^nd
4.	Academic and Industry Collaborations	Yongming Qu, Eli Lilly Weichung (Jeo) Shih, UMDNJ	Monday, June 22^nd
5.	NSF Support for Statistical Research	Professor Yazhen Wang, Univ of Connecticut	Tuesday, June 23^rd
6.	Statistical outsourcing for pharmaceutical industry: what, when and how?	Wei Shen, Eli Lilly	Tuesday, June 23^rd
7.	Conditional and unconditional exact test for contingency tables	Xin (Cindy) Wang, Pfizer	Tuesday, June 23^rd
8.	Entrepreneurship for Statisticians	Huey Lin Ju, StatPlus Inc.	Tuesday, June 23rd

Topics	Abstract	Discussion Leader
1. NIH Support for Statistical Research	Many statisticians are now working in the environment of the Department of Biostatistics at a Medical School or a School of Public Health with “soft money” support, which means that the major portion of our salary is supported by research grants. The major source of biomedical and biostatistical research grants, which are quite competitive, is the National Institutes of Health (NIH). Usually as a biostatistician, if we can provide statistical support to biomedical investigators’ research grants from which we can share a piece of the pie to support our effort and salary, it is good enough for us to survive. However, nowadays more and more academic departments of Biostatistics require their faculty to secure their own research grants as Principle Investigator (PI), instead of Statistician or Co-Investigator in biomedical investigators’ grants, in order to get promotion or some merit reward. However, with low NIH payline, it is quite challenging for biostatisticians to get our own grants. What are good strategies for us to face this challenge? There are several channels to achieve our goals, which include the general statistical methodology grant, biomedical sciences-oriented statistical grants, and statistical cores in large center or program grant applications. We will exchange ideas and share our experience on how to optimize our chance to get grants.	Hulin Wu received his PhD in Statistics from Florida State University in 1994. After two years as a visiting Assistant Professor at the University of Memphis (TN), Dr. Wu joined Frontier Science & Technology Research Foundation in 1996 as a Senior Statistician to work for Statistical Data Analysis Center (SDAC), currently Center for Biostatistics in AIDS Research (CBAR), Harvard School of Public Health. At the same time, he held an Adjunct Lecturer position at the Department of Biostatistics, Harvard University from 2000-2003. Dr. Wu joined the Department of Biostatistics and Computational Biology, University of Rochester School of Medicine and Dentistry as a full professor in 2003. He founded and serves as its Chief of Division of Biomedical Modeling and Informatics since 2005. Dr. Wu’s research focuses on modeling infectious diseases and immune response using mathematical models and statistical methods. Dr. Wu, as the PI, is currently directing the Center for Biodefense Immune Modeling with total NIH funding of about $10 million. Since 1998, Dr. Wu, as PI or Co-PI, has received the R29 FIRST award, 3 RO1 grants, one center contract, one CTSI novel methodology grant, and one CTSI Key Function grant from NIH with a total budget of about $17 million. Contact information: Hulin Wu, Ph.D., Professor, Department of Biostatistics and Computational Biology, University of Rochester School of Medicine and Dentistry, 601 Elmwood Avenue, Box 630, Rochester, New York 14642 Email: hwu@bst.rochester.edu, Tel. 585-241-0705 Website: http://www.urmc.rochester.edu/smd/biostat/people/faculty/WuSite/Index.htm
2. DNA copy number analysis of high throughput SNP/genomic array	DNA copy number variation is known to associate with development of diseases. For example, tumor cells usually undergo dramatic chromosome changes resulting in gain or loss of DNA copy numbers. High throughput array comparative genomic hybridization (aCGH) technologies have made it possible to simultaneously examine DNA copy numbers at thousands of sites of a genome. We will discuss the statistical challenges of aCGH data. Depending on the interests of participants, we will focus on topics such as quality control, batch effect, DNA segmentation, sample clustering, and signature discovery.	Ke Zhang joined Abbott Laboratories in 2007 as a senior statistician, and he is currently a senior research statistician. He received his PhD in Statistics from Kansas State University in 2008. Since joined Abbott, he has focused in genomics data analysis and biomarker discovery. His research interest includes high dimensional data, multivariate methods, longitudinal data, nonparametric methods, and Bayesian methods. Contact information: Email: kurt.zhang@abbott.com Phone: 847-937-2556 Address: 100 Abbott Park Road, AP9A-1, Abbott Park, IL 60064
3. Statistical Opportunities in Emerging Countries	In this round table I would like to discuss the capabilities and capacity that is available in India with respect to providing statistical services for the pharmaceutical industry. This discussion will primarily have as its basis, the experiences that I have personally gained in developing and growing the statistical reporting groups for two major pharmaceutical companies, Novartis and GSK. I will try to give pointers to what is needed for the operations to succeed, what kind of talent is available and how to use the talent available in a global team setting to make this work. In addition to this, various activities which are going on in India with respect to developing statistical capabilities in India would also be highlighted. This includes information on statistics societies which are active in India, academic information on how statistics education is given in India and an update on academic-industry collaboration in India.	Ashwini Mathur received his PhD in Biostatistics from University of California, Berkeley. He joined University of California, San Francisco as an Adjunct Professor in 1994 and worked there till 1997 mainly supporting clinical research in the Department of Radiology and clinical trials related to Osteoporosis. In 1997 he joined GlaxoSmithKline as a Senior Statistician in the pre-clinical safety group. He became a manager of Clinical Pharmacology group in 2000. In 2002 he moved back to India to head up GSK's Biometrics office in Bangalore, India supporting statistical reporting of Phase I-IV clinical trials. He grew the group from 10 people to 35 by 2006 when he moved to head the Clinical Information Sciences India group for Novartis. He has grown the group from 45 to close to 100 and is currently there. Besides the industry work, Ashwini is actively involved in statistics society work as Secretary for International Biometrics Society and founder member of Indian Association for Statistics in Clinical Trials. Contact Information: Ashwini Mathur, PhD Head, CIS India Novartis Healthcare Pvt . Ltd. Phone: +91 40 66576354, FAX: +91 40 66576252 Email : ashwini.mathur@novartis.com
4. Statistical collaboration between industry and academia	It has been a long history of collaboration between industry and academia. However, given that information increases, the collaboration model may need a transformation. For example, innocentive.com recently posted quite a lot of problems for scientists all over the world to solve. Eli Lilly recently started a global “Stat Network” project which aims to build a network of collaboration where Lilly is designed as one element of the network, which is different from the traditional model where Lilly is the center of its network. We will briefly review the innocentive.com and Lilly Stat Network and see how all companies and institutes can be more efficiently networked.	Yongming Qu received his PhD in Statistics from Iowa State University in 2002 and subsequently joined Lilly as a Sr. Statistician. He is currently a Principal Research Scientist at Lilly and an Adjunct Professor of Iowa State University. Since joined Lilly, he has published more than 20 articles in peer reviewed journals. His research interest includes surrogate markers, measurement error models, nonparametric methods, missing data, and analyzing non-randomized studies. Contact information: Email: quyo@lilly.com Phone: 317-6518593 Address: Lilly Corporation Center, Eli Lilly and Company, Indianapolis, IN 46285
5. Funding Opportunities at National Science Foundation (NSF)	NSF is one of major Federal Agencies to fund statistics research. The Statistics Program at Division of Mathematical Science of NSF supports research in all areas of Statistics. Besides the Statistics Program, there are many programs that support interdisciplinary research projects. During the panel discussion I will give an overview of NSF and introduce NSF programs that fund statistics related research projects.	Dr. Yazhen Wang is currently a professor of statistics, University of Connecticut. He currently serves as NSF Program Director. He received his PhD in Statistics from University of California at Berkeley in 1992. He is Fellows of IMS and ASA, and serves as Associate Editors of JASA, Statistica Sinica, The Econometrics Journal, and Journal of Korean Statistical Society.
6. Statistical outsourcing for pharmaceutical industry: what, when and how?	As an industry, pharmaceutical has experimented with various outsourcing approaches over the last 10 years. Although there have been some success stories, we are still faced with unsustainable and disparate outsourcing strategies and tactics. Recently, outsourcing of statistical work in the pharmaceutical industry has grown significantly. We will discuss and exchange ideas on a variety of issues related to statistical outsourcing in pharmaceutical industry. Our discussion encompasses current state of outsourcing, lessons learned, outsourcing model, impact of globalization, qualification and educational requirement of outsourcing workforce, roadmap to the future, etc. The goal of discussion is to achieve successful outsourcing and strong partnership in pharmaceutical R and D process that speeds innovation and meets patients’ needs.	Dr. Wei Shen received his Ph.D in biostatistics from University of Minnesota in 1996. He joined Eli Lilly and Company in 1996 as a senior statistician. During his career at Lilly, Dr Shen has provided extensive statistical support and leadership for drug development, registration and post-marketing. Currently, Dr Shen is Head of statistics, and he is responsible for leading over 40 staff members who support Phase 4 clinical trials, data mining, and health outcomes research. Dr Shen has published over 25 manuscripts in statistical methodology and medical research. Dr Shen is an elected board member of the International Chinese Statistical Association (ICSA). Contact information: Email: shen@lilly.com Phone: 317-276-8379 Address: Lilly USA LLC, Lilly Corporation Center, Indianapolis, IN 46285
7. Conditional and unconditional exact test for contingency tables	In clinical trials involving binary primary endpoint, it is often the case to compare percentage difference of responders across different treatment groups. When there are two treatment arms (treatment vs. placebo), the problem is simplified down to a two-by-two contingency table. When normal approximation is appropriate, a simple two-sample proportion test or chi-square (Pearson) test is usually adopted. When the response rate is low, exact method is preferred. It is of interest to compare some properties (type I error rate, power, confidence intervals) of chi-square test, Fisher exact test (Fisher 1922) and Barnard exact test (Suissa and Shuster 1985). Both exact tests are based on enumerating possible contingency tables, so the support of the test statistics is discrete. The measure for Fisher exact is table probability, and that for Barnard is z-score. Fisher test conditions on observed row and column totals, while Barnard conditions on column total only. Barnard exact test is recommended for two-by-two table for type I error control, increased power, equivalent confidence intervals and more intuitive interpretation. The performance of the tests for general r-by-c tables will be discussed.	Xin Wang received her PhD in Statistics from Northwestern University in 2006 and subsequently joined Sanofi-aventis as a Sr. Statistician. She is currently a Manager at Pfizer. Her research interests include multiple comparision, gatekeeping procedures and dose-response. Naitee Ting is currently a Director in the Biostatistics group of Pfizer Global Research and Development at New London, CT. Naitee received his Ph.D. in 1987 from Colorado State University (major in Statistics). He has an M.S. degree from Mississippi State University (1979, Statistics) and a B.S. degree from College of Chinese Culture (1976, Forestry). He has been with Pfizer since 1987. Contact information: Email: xin.wang@pfizer.com, naitee.ting@pfizer.com Phone: 860-7322961, 860-7324871 Address: 50 Pequot Ave, New London, CT 06320
8. Entrepreneurship for Statisticians	In this discussion, capacity required for becoming an entrepreour will be first reviewed. Possible areas for statisticians to establish entrepreneurship will be explored, followed by the discussion of keys to success and causes of failure. Case studies in marketing survey and clinical research arena and will be used as examples. Challenges experienced as a statistical consultant and running a CRO will be shared with the participants. How to strive during the economic down turn will be discussed.	HUEY LIN JU is President of StatPlus, Inc, a CRO providing services in clinical research. Ju began her career at Bristol Myers Squibb and was previously the Sr. Director of Biostatistics of ICON where she supervised clinical studies in the area of Statistics. Ju was a Visiting Scholar at Stanford University, the Founding Director of Statistical Consulting Laboratory and Associate professor of National Chengchi University, Taiwan. She was on the Advisory Committee of the Bureau of Pharmaceutical Affairs, Department of Health (DOH), an Adjunct Associate Professor of Taipei Medical University and the Founding President of Princeton-Trenton Chapter of the ASA. Contact information: Email: huey.ju@statplus.com Phone: 510-687-1483 Address: 673 Plymouth Ave, Fremont Ca, 94539