this issue as a pdf. - The International Biometric Society
Transcription
this issue as a pdf. - The International Biometric Society
BIOMETRIC Vol. 33 No. 2 • April–June 2016 BULLETIN International Biometric Society Internationale Biometrische Gesellschaft Société International de Biométrie “Biometry, the active pursuit of biological knowledge by quantitative methods.” - R.A. Fisher, 1948 President’s Corner Greetings! Although we are all looking forward to the IBC in Victoria, now only six weeks away, I am writing to you from Salvador, Brazil, where I am currently participating in the RBras 61st Annual Meeting. Salvador has one of the smaller groups of RBras members, but they have put together an excellent meeting showing the diversity of Biometry in Brazil, from medicine and public health to agriculture and the environment. Brazil of course is one of our older regions, but in other regional news, it is a pleasure to inform you that the Malawi Region is now officially formed and that this youngest region of the IBS family will host the SUSAN meeting in 2017. Here in Salvador, it has been a pleasure to meet the President of ISI, Pedro Silva. As you may know, ISI and IBS have recently formalized the relationship between the two Societies with an MOU, but this is really just recognition of our ongoing relationship as the two international statistical societies with a mission of education and research in statistical science around the world. As usual we will have an ISI Showcase Session at the IBC in Victoria, organized by Kaye Basford on the important topic of ‘Global Food Security.’ An IBS session is also planned for the July 2017 ISI meeting in Marrakech. Often at the regional meetings, it is encouraging to see so many younger members here in Salvador. Although the Meet the Presidents Roundtable had fewer participants and a higher average age than we had hoped, the thoroughly Brazilian party that followed showed there is much youthful energy here. The number of Brazilian iPhones that now contain a photo of me with some younger participant simply shows that the next generation celebrates and remembers their meetings in a different way. But for most IBS members, foremost in our thoughts is the upcoming IBC, the major biennial event of our Society, as we prepare to welcome all those who are coming to Victoria. The heaviest burden of preparation has of course fallen on the IPC and LOC, but now invited and contributed sessions are set and are being slotted. The Council’s task has been on our major awards. Three members with longtime and distinguished service to IBS have been named Honorary Life Members: Kaye Basford (Australasia Region), Peter Bauer (AustroSwiss Region) and Thomas Louis (ENAR). Additionally, the Rob Kempton Award for outstanding contributions to the development of biometry in the developing world is presented to Jane Hutton for her sustained contribution to the inclusion of biometry in the African Institute for Mathematical Sciences (AIMS) program of education in Africa.These awardees will be recognized at the upcoming IBC Awards Ceremony. At that same event, we will welcome members of the Breslow family to remember with us our former President and Honorary Life Member, Norm Breslow. The Awards Fund Committee has also been busy and has made 14 travel awards to IBC participants (five of them students) from developing countries. The recipients represent a wide geographic spread, coming from: Ethiopia, Kenya, Uganda, South Africa, Turkey, Pakistan, India and Brazil. Other Committees have been responsible for our showcase sessions. In addition to the ISI Showcase Session, we have our usual awards for Best Papers 1 Biometric Bulletin IN THIS ISSUE President’s Corner . . . . . . . . . . . . . . . . . . . 1 From the Editor. . . . . . . . . . . . . . . . . . . . . . 2 Region Key. . . . . . . . . . . . . . . . . . . . . . . . . . 2 NEW FEATURE: How much should drugs cost? . . . . . . . . . . . . . . . . . . . 4 Biometrics. . . . . . . . . . . . . . . . . . . . . . . . . . 5 JABES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6 Special Contribution of Biometrical Journal.7 New President: Elizabeth E. Thompson. . . . 7 Software Corner: Parallel computation in R. 9 Region News. . . . . . . . . . . . . . . . . . . . . . . 10 Australasian Region. . . . . . . . . . . 10 British & Irish Region. . . . . . . . . . 11 Dutch Region. . . . . . . . . . . . . . . . 11 Eastern Mediterranean Region . . 11 Eastern North American Region. . 13 German Region . . . . . . . . . . . . . . 13 Indian Region. . . . . . . . . . . . . . . . 14 Italian Region. . . . . . . . . . . . . . . . 15 Japanese Region. . . . . . . . . . . . . 15 Western North American Region. 15 Announcements and News. . . . . . . . . . . . 15 Meetings. . . . . . . . . . . . . . . . . . . . . . . . . . 16 in Biometrics and in JABES. Given the many high-quality papers in our Society’s journals, the members of the Committee that judge papers for these showcase sessions have a substantial task, and the papers they have selected promise two excellent sessions. We also have a number of sessions directed towards more junior members. The Education Committee has been responsible for the short courses, and for the first time in a number of years we have four short courses with good enrollment in each. Additionally, the very successful Statistics in Practice Session, pioneered by the German Region, will occur again; this time with two speakers from the UK on the topic of meta-analysis. Continued on p. 6 BIOMETRIC BULLETIN ISSN 8750-0434 Copyright © 2016 International Biometric Society Biometric Bulletin is published four times a year in March, June, September and December for US$40 per year by the: International Biometric Society 1444 I Street, NW, Suite 700 • Washington, DC 20005-6542 USA Telephone: +1 (202) 712-9049 • Fax: +1 (202) 216-9646 Email: <[email protected]> Website: http://www.biometricsociety.org The Biometric Bulletin is available to members of the Society as part of their annual dues. The views of contributions to this publication should not be ascribed to the International Biometric Society. Reproduction for commercial purposes is allowed if the source is acknowledged. Editor Havi Murad, Biostatistics Unit, The Gertner Institute for Epidemiology & Health Policy Research, Sheba Medical Center, Email: <HaviM@ gertner.health.gov.il> Regional Correspondents Joanna in ‘t Hout (ANed), Vanessa Cave (AR), Roland Langrock (BIR), Hein Putter (Channel Network), Axel Benner (DR), Giota Touloumi (EMR), Leslie McClure (ENAR), Mamadou Diedhiou (GEth), Olayemi Oluwasoga (GNi), Zofia Hanusz GPol, Cornelia Enachesu (GRo, Peter M. Njuho (GSAf), Wellington Mushayi GZim, Anil Mathew (IR), Satoshi Hattori (JR), Esa Läärä (NR), Cecilia Bruno (RArg), Sophie Vanbelle (RBe), Luzia Aparecida Trinca (RBras), Novie Younger (RCAC), Anabel Forte (REsp), Robert Faivre (RF), Livio Finos (RItl), Seung-Ho Kang (Rko), Dominik Heinzmann (ROeS), Henry G. Mwambi (SUSAN), Megan Othus (WNAR), Jialiang Li (SING), Dan Kajungu (GUgan), Agnes Ankomah (GGha), John Mwangi (GKe), Njoku Ama (GBot), Cristian Meza (GCl), Katja Ickstadt (CEN), Alia Sajjad (PKSTAN), Andrew Zhou (CHINA), Johny Javier Pambabay Calero (ECU) International Biometric Society Executive Board President: Elizabeth Thompson, United States Immediate Past President: John Hinde, Ireland Secretary-Treasurer: James Carpenter, United Kingdom Directors: Karen Brandeen-Roche, United States; Frank Bretz, Switzerland; Krista Fischer, Estonia; Joel Greenhouse, United States; Freedom Gumedze, South Africa; Tae Rim Lee, Korea; Sharon-Lise Normand, United States; José Pinheiro, United States; Paulo J. Ribeiro, Brazil; Alan Welsh, Australia; Ernst Wit, the Netherlands; Andreas Ziegler, Germany Editors of Biometrics Yi-Hau Chen, Academia Sinica – Institute of Statistical Science, Email: <[email protected]> Michael J. Daniels, University of Florida, Email: <mdaniels@ stat.ufl.edu> Marie Davidian (Executive Editor), North Carolina State University, Email: <[email protected]> Stijn Vansteelandt, University of Ghent, Email: <[email protected]> Editor of Journal of Agricultural, Biological and Environmental Statistics (JABES) Stephen T. Buckland, CREEM, University of St. Andrews, Email: <[email protected]> International Business Office Executive Director: Dee Ann Walker, CAE Director of Education: Alphonsus Baggett, MEd Director of Administration: Mik Bauer From the Editor Region Key Dear Readers, Regions The 28th IBC is on our door step, and I am very excited about the interesting varied program. I plan to participate in the meeting of the Committee on Communications and to hold a meeting for the Biometric Bulletin team: Regional correspondents, Software Corner correspondents, etc. on Tuesday14:00 – 15:30. Updates from these meetings along with my personal impressions from the conference will be given in the next issue. Anyone who has ideas for new corners in the Bulletin, topics for the Special Feature articles or for our NEW Software Corner can write to me or approach me in Victoria! I’ll be happy to talk to you. In this issue I am happy to introduce Daria Steigman, our professional correspondent. I figured it would be nice to know who is standing behind these interesting Special Feature articles.The article for this issue is on cost-effectiveness of drugs. The idea to write an article on this topic arose when I heard Daniel Goldstein, senior physician at the Rabin Medical Center in Tel Aviv and adjunct assistant professor at Emory University, gives a guest lecture at the Gertner Institute. He described a cost-effectiveness model to factor medication and administration costs with life expectancy, frequency and management of adverse effects and quality of life. In this model incremental survival benefit for each drug is calculated in terms of quality-adjusted life-years (QALY). In the NEW Software Corner, Garth Tarr from the Australasian Region has written an article on Parallel computation in R. I have recently programmed a Gibbs Sampling Markov Chain in SAS UNIX using a multiple imputation method for a complicated measurement error problem, and it used a CPU of 22 minutes per one simulated data. Imagine running 200 simulations…. Parallel computations could have saved me a lot of CPU time. I’ll be happy to receive your responses to this article and also ideas for future articles. As you know we regularly include in the Bulletin highlights and updates from two IBS Journals: JABES and Biometrics. Occasionally, we will include some updates from more local journals of specific regions of IBS. For example, in this issue we include a special contribution of the Biometrical Journal (German and Austro-Swiss Regions) on Reproducible Research. Editors of such local journals are invited to contact me and suggest special contributions from their journals. Biometric Bulletin 2 RArg - Argentinean Region AR - Australasian Region ROeS - Austro-Swiss Region RBe - Belgian Region GBot - Botzwanian Region RBras - Brazilian Region BIR - British and Irish Region RCAC - Central American-Caribbean Region GCI - Chilean Region CHINA - Chinese Region EMR - Eastern Mediterranean Region ENAR - Eastern North American Region ECU - Ecuadorian Region GEth – Ethiopian Region RF - French Region DR - German Region GGha - Ghanian Region IR - Indian Region RItl - Italian Region JR - Japanese Region GKe – Kenyan Region RKo - Korean Region GMal - Malawi Region GNi – Nigerian Region NR - Nordic-Baltic Region PKSTAN – Pakistani Region GPol – Polish Region GRo – Romanian Region SING – Singaporean Region GSaf – South African Region REsp - Spanish Region ANed - The Netherlands Region GUgan – Ugandan Region WNAR - Western North American Region GZim – Zimbabwean Region Networks CEN - Central European Network CN - Channel Network SUSAN - Sub-Saharan Network I would like to thank the outgoing president of IBS, John Hinde. John attended the EMR conference in Cappadocia Turkey during my presidency, and I appreciate his good advice and contribution to IBS. John has kindly agreed to interview the new President, Elizabeth E. Thompson, for this issue. I will finish up by mentioning a “Statement on Statistical Significance and P-Values” with six principles underlying the proper use and interpretation of the p-value, released by the American Statistical Association (ASA) last February in TAS. Although these issues have already been discussed, this is the first time that a community of statisticians as represented by ASA has issued a statement to address these issues. The statement’s six principles are: 1.P-values can indicate how incompatible the data are with a specified statistical model. 2.P-values do not measure the probability that the studied hypothesis is true, or the probability that the data were produced by random chance alone. 3.Scientific conclusions and business or policy decisions should not be based only on whether a p-value passes a specific threshold. 4.Proper inference requires full reporting and transparency. 5.A p-value, or statistical significance, does not measure the size of an effect or the importance of a result. Selected comments of IBS members on the statement (all 22 comments appear in the Supplemental of the original statement): 6.By itself, a p-value does not provide a good measure of evidence regarding a model or hypothesis. 1.Yoav Benjamini (EMR): http://errorstatistics. files.wordpress.com/2016/03/benjamini.pdf. Original statement: http://amstat.tandfonline. com/doi/abs/10.1080/00031305.2016.1154 108. A brief outline: http://www.amstat.org/newsroom/pressreleases/P-ValueStatement.pdf. 2.Roderick J. Little (ENAR): https://s3-euwest-1.amazonaws.com/pstorage-tf-iopjsd8797887/4858042/15_Little.pdf. 3.Stephen Senn (Austro-Swiss): https://s3-euwest-1.amazonaws.com/pstorage-tf-iopjsd8797887/4858054/19_Senn.pdf. Havi Murad XXVIIIth International Biometric Conference IBC 2016 2016 INTERNATIONAL BIOMETRIC CONFERENCE July 10-15, 2016 Victoria, British Columbia, Canada The 2016 International Biometric Conference will feature a wide array of exciting topics, including: Biomarkers Complex study designs Correlated data Electronic medical records Fisheries Forensic statistics Genomics and other omics fields Human rights Network analysis Neuroimaging 3 Nonparametric and semiparametric statistics Missing data Observational data and causal inference Pharmarcokinetics and pharmacodynamics Reproducible research Spatio-temporal statistics Statistics in science Biometric Bulletin Meet Our Professional Correspondent, Daria Steigman! Daria Steigman is an entrepreneur, writer, and founder of marketing / PR consultancy Steigman Communications. As a contributor to Biometric Bulletin, she is charged with taking research and explaining how its application matters to the broader public and policy conversations around public health, the environment, and other bioscience disciplines. New Feature: How Much Should Drugs Cost? Exploring the cost of prescription drugs – and their value to patients. By Daria Steigman Anyone who pays retail for health care in the United States knows that there is no set fee for service. There are prices negotiated by individual insurers, the price negotiated under Medicare, and what each provider would love to charge in a perfect world. That’s the high price of retail—for people with giant deductibles and those without insurance. Sometimes you can “shop” for a service (or negotiate a “discount” later), but most of the time the retail patient is just a powerless buyer in a seller’s market. The arbitrary price of health care services came into sharp focus for many Americans last fall when Martin Shkreli, aka “the pharma bro,” overnight raised the price of Daraprim (used to treat toxoplasmosis) from $13.50 to $750 per pill. At the time, Shkreli told CBS News* that “at this price it’s a reasonable profit. Not excessive at all.” Which raises the question: how much should prescription drugs costs? It’s not an easy question. Just start talking about cost and value, and you have some people thinking that it’s all about how much it’s “worth it” to keep someone alive. On the flip side, patients are spending billions of dollars on drug treatments that may or may not prolong live—with often-devastating side effects that massively reduce patients’ quality of live. To understand the question, Biometric Bulletin talked with two researchers who are exploring the pricing and cost-effectiveness of oncology drugs. A Framework for Value-based Pricing Daniel Goldstein, senior physician at the Rabin Medical Center in Tel Aviv and adjunct assistant professor at Emory University, has been studying cancer drugs. He told Biometric Bulletin that the prices have skyrocketed—without any link to their effectiveness. Goldstein says that the drug development and approval process in the U.S. is missing a key step: the evaluation of cost and value. To address this missing link, he and his fellow researchers at Emory University created a framework for establishing a value-based price for new cancer drugs entering the market. It is a cost-effectiveness model that factors in medication, administrative costs, life expectancy, frequency and management of adverse effects, and quality of life. Goldstein says that physicians are making a mistake when they don’t take cost into account. “This is an issue for doctors,” he says. “We have to be sensitive to the patient’s right, and the cost and benefit when one is a lot higher.” While the current model is focused on the U.S. market, Goldstein and his colleagues are now turning their attention to creating models that look at the cost of drugs in different countries (including Canada, Australia, England, Israel, South Africa, India, and China). Comparing Treatment Options While Daniel Goldstein is looking to build a broad framework, Simon Zeichner looked specifically at the cost effectiveness of three immune therapies used in the treatment of metastatic melanoma (ipilimumab, nivolumab, and pembrolizumab). The study looked at the survival curves from the previously published phase 3 trials that included patients with metastatic melanoma and compared the different sequencing of these drugs. Zeichner told Biometric Bulletin that many of the new immune therapies are associated with an unprecedented number of patients experiencing severe immune-related adverse effects. To understand the options and many variables involved in such an analysis, Zeichner and his colleagues built a comprehensive software model that allowed them to input the costs of all aspects of oncologic care (e.g., infusions, doctor visits, drugs, treatment of adverse effects, hospitalizations). In order to gauge patient quality of life, they also incorporated previously published health utility values (from patient survey data) for each disease state and adverse effect. The model is, says Zeichner, “one of the most comprehensive cost-effectiveness models to date.” Take ipilimumab, a CTLA-4 inhibitor, as an example. The drug was recently approved by the U.S. Food and Drug Administration (FDA) to prevent disease recurrence in patients with locally advanced melanoma that has been completely surgically removed, but at a dose (three times as much for a 70 kg. patient) and treat- Biometric Bulletin 4 ment period (three years vs. three months) that far exceeds what is FDA-approved for use in patients with metastatic melanoma. FDA approval compared ipilimumab to a placebo rather than to interferon alpha 2b, the previous standard of care. Zeichner’s model demonstrates that the higher dose is associated with a significantly greater cost (estimated at more than $1.4 million for the drug alone—not counting doctor visits, scans, infusion costs, treatment of adverse effects, or hospitalizations). It also suggests that the lower dose is not only equally effective but also better tolerated by patients. More important, both Zeichner’s and Goldstein’s research are about getting the best “bang for the buck” when it comes to health care spending. While pharmaceutical companies are looking for ways to maximize their profits, the heavy lifting is left to health economics researchers to develop cost-effectiveness models to understand the “right” cost and value—and the best and most cost-effective ways to treat patients for the least amount of toxicity, cost, and more. That’s where science and statistics can help insurers and policymakers frame the debate for what is, inevitable, always going to be a tough conversation. In addition, Zeichner’s model demonstrates that the higher dose is associated with significant toxicity. During the arbitrary threeyear treatment period used in the phase 3 trial, more than half of the patients who received ipilimumab had to discontinue treatment as a result of adverse effects (some of which were associated with significant morbidity and required lifelong treatment). Zeichner added that the study that formed the basis for *CBS News citation: http://www.cbsnews.com/news/turing-pharmaceuticals-ceo-martin-shkreli-defends-5000-percent-price-hike-ondaraprim-drug/ Biometrics September 2016 Issue Highlights Papers in the September Biometric Methodology section include “A flexible ratio regression approach for zero truncated capture-recapture counts,” by Dankmar Boehning, Irene Rocchetti, Marco Alfo, and Heinz Holling; “Spatial regression with covariate measurement error: A semi-parametric approach,” by Md Hamidul Huque, Howard D. Bondell, Raymond J. Carroll, and Louise M. Ryan; “Model selection and inference for censored lifetime medical expenditures,” by Brent A. Johnson, Qi Long, Yijian Huang, Kari Chansky, and Mary Redman; “An information-theoretic approach for the evaluation of surrogate endpoints based on causal inference,” by Ariel Alonso, Wim Van der Elst, Geert Molenberghs, Marc Buyse, and Tomasz Burzykowski; and “Interpretable functional principal component analysis,” by Zhenhua Lin, Liangliang Wang, and Jiguo Cao. The Biometric Practice section features papers spanning a diverse range of application areas, including “Model assessment in dynamic treatment regimen estimation via double robustness,” by Michael P. Wallace, Erica E. M. Moodie, and David A. Stephens; “Estimation and interpretation of heterogeneous vaccine efficacy against recurrent infections,” by Juha Mehtala, Ron Dagan, and Kari Auranen; “Unbiased estimation of biomarker panel performance when combining training and testing data in a group sequential design,” by Nabihah Tayob, Kim-Anh Do, and Ziding Feng; “Global rank tests for multiple, possibly censored, outcomes,” by R. Ramchandani, D. A. Schoenfeld, and D. M. Finkelstein; and “An efficient design strategy for logistic regression using outcome- and covariate-dependent pooling of biospecimens prior to assay,” by Robert H. Lyles, Emily M. Mitchell, Clarice R. Weinberg, David M. Umbach, and Enrique F. Schisterman. As always, lists of papers to appear can be found at the Biometrics website. Papers to appear in future issues may also be found under the “Early View” link at the Wiley website, which may be accessed by IBS members by visiting http://www.biometricsociety.org/, selecting “Biometrics” from the drop-down menu at the “Publications” link at the top of the page, and accessing the “Click here” link. Biometrics Showcase Session at IBC2016 The Biometrics Showcase Session at the IBC in Victoria will be held Tuesday, 12 July, 4:00 pm - 5:30 pm. As reported in a previous col5 umn, the session will feature presentations of the two winners of the Best Paper in Biometrics by an IBS Member Awards for 2014 and 2015, “Bayesian nowcasting during the STEC O104:H4 outbreak in Germany,” by Michael Hohle and Matthias an der Heiden, published in the December 2014 issue, presented by Dr. Hohle; and “Handling missing data in matched case-control studies using multiple imputation,” by Shaun R. Seaman and Ruth H.Keogh, published in the December 2015 issue, presented by Dr. Keogh. The authors will receive the awards at the IBC Award Ceremony that same evening. Editorial Board News Co-editor Yi-Hau Chen’s term will end 31 December 2016. According to geographic convention, the new Co-editor (CE) should reside outside of Europe or North America. A Search Committee has been appointed according to IBS Policies and Procedures, which state that the Committee should include the Chair of the Editorial Advisory Committee (EAC), two members of the EAC, the Biometrics Executive Editor, and the current Biometrics Co-editors. The committee members are: Marie Davidian, Biometrics Executive Editor, Chair (ENAR); Esa Läärä, EAC Chair (Nordic Baltic Region); Liliana Lopez Kleine, EAC member (Central America-Caribbean Region); Julio M. Singer, EAC member (Brazilian Region); Yi-Hau Chen, Biometrics CE (Chinese Region/At-Large); Stijn Vansteelandt, Biometrics CE (Belgian Region); and Mike Daniels, Biometrics CE (ENAR). The committee has identified a nominee; the new CE will be announced in a future column upon approval of the nomination by the IBS Executive Board. News from Our Publisher Through an initiative of our publisher, Wiley, Biometrics will begin participating in the peer-review recognition program Publons. Publons (https://publons.com/) works with reviewers, publishers, universities, and funding agencies to turn peer review into a measurable research output. Specifically, Publons collects peer-review information from reviewers and from publishers and produces comprehensive reviewer profiles with publisher-verified peer-review contributions that referees can include in their promotion and funding applications. In Fall 2015, Wiley piloted Publons with several Wiley journals and received positive feedback from referees. Wiley has now offered this opportunity to all of its society-owned publications. Biometric Bulletin Reviewers for Biometrics will be able to opt to use Publons to track and verify their peer review activity through our ScholarOne editorial management system. We welcome feedback from our referees on this program. Please send comments to [email protected]. Wiley has provided the IBS with its 2015 Publisher’s Report for Biometrics. Some highlights: Over 5,200 institutions purchased access to Biometrics worldwide in 2015 through either a Wiley license or traditional single-title subscription, compared with about 4,600 in 2014. Over 5,400 institutions in the developing world now have access to Biometrics through free or low-cost subscriptions. Article downloads increased by 4% over 2014, and the average number of downloads per article was 44, with some articles being downloaded over 1,000 times. Over 2,700 individuals worldwide are registered to receive automatic alerts of new Biometrics content. Journal of Agricultural, Biological, and Environmental Statistics (JABES) Editor Report We continue to see a healthy number of papers coming through the system. Following 12 papers in the March issue, a further 10 papers appeared in the June issue. The topics of these articles were: “Nonresolvable Row–Column Designs with an Even Distribution of Treatment Replications”; “A Comparison of SuperValid Restricted and Row–Column Randomization”; “Estimability Analysis and Optimal Design in Dynamic Multi-scale Models of Cardiac Electrophysiology”; “A Nonlinear Mixed-Effects Model for Multivariate Longitudinal Data with Dropout with Application to HIV Disease Dynamics”; “Randomly Truncated Nonlinear Mixed-Effects Models”; “A Bayesian Superpopulation Approach to Inference for Finite Populations Based on Imperfect Diagnostic Outcomes”; “The Micro-behavioral Framework for Estimating Total Damage of Global Warming on Natural Resource Enterprises with Full Adaptations”; “Categorising Count Data into Ordinal Responses with Application to Ecological Communities”; “Using Species Proportions to Quantify Turnover in Biodiversity”; and “Measuring the Inspectorate: Point and Interval Estimates for Performance Indicators”. The Special Issue on “Space-time Analysis of Natural or Anthropogenic Catastrophes”, edited by Jorge Mateu and Emilio Porcu, is nearing completion. For more information on upcoming issues, the editorial board, and the aim and scope of the journal, please visit our website http://link. springer.com/journal/13253. We also accept submissions of books to review in the upcoming issues of JABES; to submit a book for review, please see the above website (click on “Editorial Board”), or contact Ken Newman ([email protected]). Steve Buckland Editor in Chief President’s Corner Continued from p. 1 Last but not least, we have our Young Statisticians’ Showcase Session. Yet another committee had the hard task of selecting five papers out of more than 50 submitted. The five winners who will give their oral presentations are chosen to represent the five major continental regions, so most will travel a long way to Victoria. So that they and other younger members can not only enjoy their week in Victoria but also perhaps make more lasting contacts with each other, yet another team of participants (some recruited from among younger registrants) are setting up communication links and preparing activities directed towards the younger generation. As we are all well aware, it is on involvement of the next generation that the ongoing health of our Society ultimately depends. And finally our local hosting Region, WNAR, has combined its own annual meeting into the IBC. WNAR has its own exciting schedule of sessions, including the Young Statisticians’ Papers Competition with 23 competitors! So their members have been doubly busy, not only with the LOC but also with their own meeting arrangements. So it is clear that many of you have been involved, at many different levels, and to all who have contributed, I give my sincere thanks. We are a volunteer Society and dependent on the commitment and enthusiasm of our members. Thanks go also to all who organized sessions, to those who will speak in those sessions, to others who are preparing their oral presentations or posters and to those who will come and participate in any way. Together we can look forward to another successful and thoroughly enjoyable IBC. Elizabeth Thompson The Invited Session at this year’s International Biometric Conference in Victoria, Canada will feature two papers that won best paper awards. For 2014, the award went to Gurutzeta Guillera-Arroita, Martin Ridout and Byron Morgan for their paper on “Two-stage Bayesian Study Design for Species Occupancy Estimation”, and the 2015 award winners were Gustavo de los Campos,Yogasudha Veturi, Ana Vazquez, Christina Lehermeier and Paulino Pérez-Rodriguez, whose paper on “Incorporating Genetic Heterogeneity in Whole-genome Regressions using Interactions” appeared in the genomics special issue. We are keen to publish papers that summarize the state of methodological development in subject areas for which technological advances are generating a demand for new statistical approaches. If such papers also speculate on likely future developments, so much the better. If you feel that you could offer such a paper or can suggest a topic together with possible authors, I would be very pleased to hear from you. I would also welcome suggestions of topics for which change is sufficient to merit a set of papers suitable for a special issue. Biometric Bulletin 6 A Special Contribution of the Biometrical Journal (Published by the German & Austro-Swiss Regions) Reproducible Research At the 61st Biometrisches Kolloquium, last year’s annual meeting of the German Region (DR) organized by the Faculty of Statistics, Technical University Dortmund, a tutorial on ‘Reproducible Research’ was held. The course aimed at giving insight into the role of ‘Reproducible Research’ for biometry and the work of biometricians, to communicate the principles of reproducibility in applied biometrical research and to teach the participants literate programming, including Sweave und knitr embedded in RStudio.This included training in project management software such as Subversion (SVN) and Git/GitHub (http:// benjaminhofner.de/talks.html). ‘Reproducible Research’ has been a theme of the DR for some time; the ‘Reproducible Research’ initiative accounts for the paradigm that scientific results are incomplete, if not useless, if they cannot be reproduced and replicated. Regarding biometrical science, the tutorial in 2015 closely linked that paradigm with the replication of biometrical data analyses, both, when statistical methods are applied to real data, as well as when they are examined in computer experiments and simulation studies. As a consequence publications on biometrical methodology and biometrical practice in life sciences including medicine, environmental sciences and agriculture should be accompanied by all relevant material to reproduce the results and findings such that researchers can verify those and build upon them.This request does have a direct impact on the publication policy of journals publishing in biometry and biostatistics. Most journals support now reproducible research by providing data and program codes as supplementary material. However, ways of submission, degree of examination of that material and details of communicating those in publications reporting differ widely. to ‘Reproducible Research’ since 2009, see Hothorn T., Held L. and Friede T. (Biometrical Journal 51, 553-555).Thereby, authors are strongly encouraged to submit computer code and data sets as used in their manuscript. After verification that these materials reproduce the presented analysis, these will be published as Supporting Information on the journal’s webpage once the paper has been accepted for publication. Recently, B. Hofner, M. Schmid and L. Edler (2016) published a review and guidelines on ‘Reproducible Research’ in statistics in the Biometrical Journal (DOI:10.1002/bimj.201500156), with special consideration of the current practice used for the journal.They reviewed the experience of the past six years realizing that the number of manuscripts with ‘Reproducible Research’ material has increased by a factor of about four since 2009. However, they also noted that the code submitted as Supporting Information was of insufficient quality in more than half of the papers at first place. Often the code could not immediately be executed, and prominent issues were missing, incomplete, and/or erroneous code, failing to reproduce all tables and figures of the paper and, to a less extent, also the programming style. Striking was also the high portion of submissions missing a README file. Therefore, Hofner et al. (2016) provide information which should equip authors with a comprehensive guidance to comply with ‘Reproducible Research’ when submitting papers to a biometric or biostatistical journal. The specific guidance for the Biometrical Journal is available at the journal’s website (http://onlinelibrary.wiley.com/journal/10.1002/%28ISSN%291521-4036/homepage/ RR_Guideline.pdf). While it is used by the ‘Reproducible Research’ Editor and his team who check each paper after acceptance, this guidance together with the review paper might help to propagate ‘Reproducible Research’ in general. The Biometrical Journal, http://www.wiley-vch.de/publish/en/journals/ alphabeticIndex/2221/, published by Wiley in cooperation with the German and the Austro-Swiss Region of the IBS has committed itself Benjamin Hofner, Erlangen; Matthias Schmid, Bonn; Lutz Edler, Heidelberg IBS New President: Elizabeth Thompson Interview by John Hinde (Outgoing President) John: Elizabeth, congratulations on your appointment as President of the International Biometrics Society. We can talk more about the IBS in a moment, but perhaps first we can discuss how you became interested in statistics. I notice that your first degree, like mine, is in mathematics, so what attracted you to statistics and how did you get started? Elizabeth: Well, I come from a family of mathematicians and so mathematics was the natural thing to do.There was almost no statistics in the Cambridge math trips in the late 60’s, but I enjoyed what there was, especially the probability models and stochastic processes. But, I also had a general interest in genetics from a very early age, almost as young as I can remember. For example, I was interested in coat colors of black and yellow Labrador dogs, which we had, and during high school I did a project where I first encountered estimating allele frequencies for recessive traits. In addition, the fact that my mother, whose parents were second cousins, has a very rare recessive condition also attracted me to genetics. So I saw statistics as a way to combine mathematics and genetics.The Genetics was always the driving force to do Statistics. 7 Members of the new IBS Executive Board meet in Nijmegen, the Netherlands (April 2015). Elizabeth Thompson and John Hinde are standing in the middle of the front row. Biometric Bulletin What about the future? What do you see as exciting, productive areas of research in the next few years? In genetics, the data have changed dramatically in the last 45 years, from working with blood and enzyme markers to DNA markers that gave us genetic maps (first huge change happened in 1980) and now we have DNA sequence data in huge amounts. . So, although the basic questions are still the same – what are the genes, what do they do, and how do they do it – we are now far less dependent on the traditional genetic models and data structures and we can answer questions from population data that we could only dreamt about a few years ago. This change is going to continue in the future, and therefore we are going to need new approaches to address new types of data. That is certainly true. Now most of your life you have had a teaching position as well. How have your teaching, consulting, collaboration and research linked together and influenced one another? The combination of teaching and research is a big part of why I moved to the States in the mid-80s. I actually did a post-doc in genetics at Stanford and after that I was on the math faculty in Cambridge for 10 years, but all my research collaborations were still in the States. So, for 10 years, there really wasn’t a connection between my teaching and research. From its founding here, the University of Washington Department of Statistics was a leader in genuine statistical science collaborations and so when the opportunity came up I made the move and the ability to combine the teaching and research was a big motivation – one that worked out well for me. That leads me nicely to what I was going to ask next, on how the climate may have changed for postgraduates and faculty starting out today. Clearly, your putting together of teaching and research was not entirely straightforward, but do you think that it is more difficult for students today? I think that getting into a field like genetics and other scientific fields is much harder today than it was in the past, as there is so much more stuff out there, so much more background.Those of us who grew up with a field and learned it as it developed, as I did with statistical genetics, we were really very, very fortunate. The huge mass of literature and science that’s out there now is very daunting, I think, to students getting into any field. Following on from that, do you have any advice for students considering a career in statistics? How they might think about approaching it and getting into some area? Students often ask whether there is a particularly good area, or good topic, to work on from the point of view of their career and I always find that very hard to answer, because really my only advice is that you have to work on the problems that interest you. Nothing else is really going to work out, you can’t put your whole mind and energy into what doesn’t fascinate you. In research, thinking of the right questions, knowing what the interesting questions are is key, is rather more important than thinking of the answers. The answers will come if you have the right questions. So, you really do need to be curious about what you are working on and then the questions will come. Talking to the scientists in the area that really does interest you is probably the way to go. For me that’s what worked. Have there been people or events that have been particularly influential in your career, suggested lines of research, or whatever? Scientifically, first of all my PhD advisor Anthony Edwards, who really got me into statistical genetics, but not only for the statistical genetics but also for his interest in the foundations of statistical inference, which I know very much influenced my own thinking over the years. And others, really they’ve been the statistical and mathematical population geneticists that I’ve worked with - more senior people in the States, Jim Neel of the University of Michigan and James Crow of the University of Wisconsin, to name two. I first visited those two departments in 1975, while I was a post-doc at Stanford, and both those visits really were transformative events in the way that I then went on to do things. Turning now to the Biometric Society, when did you first become a member and why? Was there any particular reason that you felt it a natural home to join? I think that it was probably in 1972, or 73, and it was for the same reason that most students join societies, that my advisor, Anthony Edwards, told me that I should. But it was a very natural thing to do. One of my very first papers was a note in Biometrics published in 1972 and my 1974 Biometrics paper is one that is still cited. So the Biometric Society really felt a very natural home to be in. You have been a member for a good number of years, over this time what has membership of the Society meant to you? In terms of the scientific society, it is really the society that’s been everything to me. It’s the one society that I have always identified with and participated in, ever since I was a graduate student. The Biometric Society has always been my primary affiliation, because of its broad biological scientific focus, which again seems to correspond with my thinking. I have always found the members of the Biometric Society and their scientific interests so much more interesting to me than those of the straight statistical societies. But being a society of this nature, with its link between biology and statistics, isn’t without some difficulties. Do you see any great challenges that face the Society in the coming years? I think, as with all scientific societies, it’s, paradoxically, harder in this digital age to maintain the same community and communication that we used to do. One would think that the internet and digital communication and the universal availability of scientific material, they are all good things and one would hope that they would make life easier, but, in fact, in a way they make it harder to hold societies together. The younger generation of scientists has grown up in a very different environment and they don’t see the conferences and those individual interactions in the same way. On top of that they are very busy, with many, many pressures, and therefore they do not see the benefits of getting involved in the Society. You’ve alluded to it already, in terms of this linkage between biology and statistics. Do you see any other great strengths of virtues of the IBS? Yes, well the disciplinary diversity is an enormous strength, but the second is our geographic diversity. We are truly an international society with a mission to enhance biometric science around the world and especially in developing countries. So I really think that the strength of the Society is in its regions, which vary so widely in size, in focus, and in resources. Let me finish by exploring what you hope to do during your presidency? What will be your aims and objectives over the next year, or so, for benefitting the Society and the biometrics community? I think that those two things that I have mentioned are places where I would really focus. First, the regions, to get even better communication and collaboration within and among the regions and I think our Networks Program, which the Society has started, can really help some of the smaller regions to get together and perhaps have more impact. I think that visits to the regional meetings by leading members of the Society are something that we really want to do. That may help younger people in the Biometric Bulletin 8 regions get enthusiastic, not only about their own region, but about the broader Society. Short courses and other education events in the developing countries are also a very big part of that. So, that’s the regional and geographic side, and then I hope to try, and I know that this has been something that other Presidents have also worked on, to maintain the scientific diversity and the balance of the science in the society. Sometimes, the better funded clinical, epidemiological, health sides seem to dominate, in some of the regions at least. Agriculture is also key in many parts of the world and I think that ecology and environment are important to all of us and I think that the Biometric Society is what it is because of its scientific diversity. We gain so much by keeping a broad scientific perspective. Elizabeth, thanks very much for your time, the interesting discussions that we have had and your vision for the Society. All that leaves me to say is that I hope that you enjoy your term of office as much as I did mine. Thank you John and thanks for all of the advice that you have given me as I start out on this road. Software Corner Parallel computation in R Garth Tarr Parallel computing refers to situations where calculations are carried out simultaneously, for example distributing the calculations across multiple cores of your computer’s processor, as opposed to having the calculations run sequentially on a single core. Parallel computing is particularly suitable for ‘single program, multiple data’ problems, for example in simulations and bootstrapping. Parallel computation in R has come a long way over the last 10 years. If you tried to parallelise your R code a few years ago, you probably worked with the architecture specific snow (Windows) or multicore (Unix-like) packages. Since 2011 R has supported parallel computation as part of the base distribution with introduction of the parallel package (R version 2.14.0 released in October 2011). The parallel package builds on multicore and snow to provide a (mostly) platform agnostic method of leveraging multiple cores to speed up the computation of embarrassingly parallel problems. This note discusses how to use parallel and associated packages, with little or no additional effort on the part of the statistical practitioner, to speed up data processing and statistical analysis pipelines. Getting Started The parallel package is part of base R which means that it’s already installed and you can’t find it on CRAN. You can load it in the usual way library(“parallel”). The first thing you’ll want to do is detectCores() which checks how many cores you have available wherever R is running (probably your laptop or desktop computer). Parallel Apply The family of apply functions (apply, lapply, tapply, sapply, etc.) in R provide an extremely convenient way of applying a function to the margins of an array, matrix, list, etc. If you’re using apply on a data frame, e.g. apply(X,2,median) to compute the median of the columns of X, you should consider using lapply instead because it’s much faster, e.g. lapply(X,median) which computes the median of each variable in the data frame X (there’s no need for a margin argument). The apply functions operate serially, i.e. they calculate each value in sequence and are constrained to use only one of your computer’s cores. 9 If you’re on a Unix-like system, the mclapply function is an easy way to distribute the computation across the available cores. To use mclapply, first code your calculations as a function that can be called with lapply. Make sure it works serially using lapply and then use mclapply to perform the computations in parallel. library(«parallel») X <- data.frame(matrix(rnorm(1e+07), ncol = 200)) mclapply(X, median) In this simple example, there is only a relatively minor speed improvement. To get the most out of parallel processes, the functions to be run in parallel should be non-trivial. There is an overhead associated with forking the process, so it is possible to slow your code down with mclapply if the time taken to send the tasks out to various cores takes longer than performing the task serially. Windows can’t “fork” processes in the same way that Unix-like systems do. This means that you won’t see any improvements in speed when using mclapply on a Windows machine. On the plus side, your code won’t break - it will work as if you were simply using lapply. Microsoft has announced plans to incorporate a Ubuntu image into future releases of Windows 10 through a new infrastructure they’re calling “Windows Subsystem for Linux”. This means you may soon be able to run R in a native Unix-like environment (which supports forking and hence mclapply) on Windows 10 machines. Parallel Loops An alternative to mclapply is the foreach function which is a little more involved, but works on Windows and Unix-like systems, and allows you to use a loop structure rather than an apply structure. To use foreach you need to register a “parallel backend”, for example using the doParallel package. The doParallel package acts as an interface between foreach and the parallel package. A simple example of how this works is given below where we calculate a percentile bootstrap confidence interval for a least absolute deviations (LAD) regression parameter. In the code below, we resample a dataset 10,000 times and each time generate LAD regression coefficients. library(«quantreg») # for quantile regression function rq() data(engel) # the data set we’ll use # help(engel) # plot(foodexp ~ income, data = engel) # fit1 <- rq(foodexp ~ income, tau = 0.5, data = engel) # abline(fit1) library(«foreach») library(«doParallel») cl <- makeCluster(2) # create a cluster with 2 cores registerDoParallel(cl) # register the cluster res = foreach(i = 1:10000, .combine = «rbind», .packages = «quantreg») %dopar% { # generate a bootstrap sample boot_dat <- engel[sample(1:nrow(engel), replace = TRUE), ] # fit the model fit1 <- rq(foodexp ~ income, tau = 0.5, data = boot_dat) # return the coefficients fit1$coef } stopCluster(cl) # shut down the cluster In the foreach() function we’ve specified that the results should be combined using rbind (i.e. the rows will be bound together, where the ith row is the LAD regression coefficients from the ith bootstrap Biometric Bulletin sample) and we’ve indicated that the quantreg package needs to be loaded on each of the processes. The output is a matrix, that we’ve called res, which consists of two columns and 10,000 rows. We obtain a 95% percentile bootstrap by extracting the appropriate quantiles: resdf <- as.data.frame(res) quantile(resdf$income, probs = c(0.025,0.975), type = 1) ## 2.5% 97.5% ## 0.4679890 0.6137351 The for loop syntax is very similar to a regular for loop in R, except it starts with foreach and uses the %dopar% function. If you use %do% instead of %dopar% the loop is evaluated sequentially where one CPU will run at 100% until the job is finished. In the above example we asked for two cores to be used, hence with %dopar% both cores will run at 100% until the job is done. This means that a %dopar% loop running on two cores will finish in roughly half the time that it would have taken to run on a single core. As an example, the code above with 100,000 replications using %do% takes 3.0 minutes on my computer whereas using %dopar% takes only 1.6 minutes. It’s not exactly two times faster because of the computational overheads involved in sending the tasks out to the different cores and collating the results. For more details see the doParallel and foreach vignettes. Distributed Computing The function mclapply can only use the cores of one machine, i.e. jobs can’t be distributed over several nodes of compute cluster. One way to do this in R is to use the parLapply function which can utilise the Message Passing Interface (MPI) system. For further details see the Rmpi package. The doMPI package provides an MPI parallel backend for the foreach package. Other Considerations Parallel random number generation: When bootstrapping or performing simulation studies, it is desirable for each instance of R to generate independent, reproducible pseudo-random numbers. If there is an object .Random.seed in your R workspace that is then shared with the worker cores, all your instances of R may inavertently run identical simulations using identical “random” numbers. Alternatively, if .Random.seed is not in the workspace, then you will have independent streams of random numbers but it will not be reproducible.The parallel package includes a random number generator designed to overcome these issues. It can be enabled using RNGkind(“L’Ecuyer-CMRG”). See the documentation of the parallel package for details. Bioconductor: If you work with Bioconductor packages, you should look into BiocParallel which provides modified versions of functions optimised for parallel evaluation, tailored for use with Bioconductor objects. So the next time you think about leaving a simulation running on your computer for the weekend, consider using mclapply instead of lapply or rewriting that for loop as a foreach loop and have it run overnight instead, or send it out to a supercomputer and have the results within a couple of hours! 29 April 2016 Region News Australasian Region (AR) Alesha Hatton Alesha is currently an honors student at the University of Wollongong, where she completed a Bachelor of Medical Mathematics (Dean’s Scholar), graduating with distinction. For her honors year, Alesha is completing a project of the analysis of age and growth of endangered, juvenile green turtles using mixed models at the National Institute for Applied Statistics Research Australia (NIASRA). IBS-AR Student Scholarships To help attract enthusiastic and talented students to career paths in biometrics, the Australasian Region offers scholarships for suitably qualified students who intend to undertake a fourth or honors year of study, or a coursework Masters, in statistics, mathematical statistics, biostatistics, bioinformatics or biometrics. We are delighted to announce the winners of this year’s highly contested award: Clayton D’Ath (University of Waikato), Alesha Hatton (University of Wollongong) and Amanda Wright (University of Sydney). Congratulations! Clayton D’Ath Clayton is in the last semester of his undergraduate degree in statistics after which he will begin his honors degree on a yet to be decided project. Further post-graduate study in a Masters or PhD is becoming a greater reality for him with the subtle encouragements from his supervisor in the form of emails containing links to PhD applications. Over the past summer he had a studentship with AgResearch Ltd (New Zealand) where he performed exploratory data analysis on GeoChip data for the 50 Pastures Project (http://www.soilmicronz.net/50-pastures-project.html). In his personal time he is an avid hiker and takes advantage of his height in netball and ultimate frisbee. During January she attended the 2016 Australian Mathematical Science Institute (AMSI) Summer School at RMIT University, completing a subject in Complex Networks which had a largely practical emphasis on analyzing network data from epidemiology and ecology, using mathematical and statistical techniques. In 2015 Alesha attended the AMSI BioInfo Summer, a bioinformatics conference giving exposure to new areas of research in epigenetics, translational genomics and proteomics. This is a dramatically expanding area which Alesha hopes she can get into when she finishes her honors. Biometric Bulletin Amanda Wright Amanda completed her undergraduate degree at the University of Sydney in Animal and Veterinary Bioscience in 2009 and was awarded first class honors, investigating gene regulation in a neurological disorder in dogs. Following this, she began her PhD at the Garvan 10 Institute of Medical Research in 2010 to investigate molecular mechanisms of cell loss in an Alzheimer’s disease model and was awarded her PhD in 2014. She currently works as a Research Officer at the Garvan Institute in order to investigate mechanisms of neurodegeneration in a variety of disease models. During her research career Alesha fundamentally believed that having a stronger understanding of statistical design and analysis could strengthen her scientific research, so she began a Masters of Medical Statistics at the University of Newcastle in 2015. Alesha now aims to complete her masters in order to consult scientists in statistical concepts and research design and to carry out robust and reliable medical research. for Cochrane. Richard described random-effects meta-analysis in the modern era and through an amusing analogy with Homer Simpson trying to jog, posed the question “Are we running before we can walk?” Theo discussed more sophisticated methodology involving generalized linear mixed models for random effects meta-analysis but also said, “There is still a place for D&L methodology, since it works well in larger meta-analyses and for any choice of effect measure.” Finally Nan Laird led a short discussion about “How heterogeneity can be a good thing” and gave an amusing example of a sign from New Cuyama, where the population (562), ft above sea level (2150) and “Established” (1951) were added to give 4663. This was a highly heterogeneous dataset if ever there was one! IBS-AR Travel Awards The Australasian Region offer travel awards to assist outstanding students and/or early career members to attend international meetings of the society. We congratulate Garth Tarr, a lecture in statistics at the University of Newcastle and Correspondent of the Software Corner of the Biometric Bulletin, and Lee Yan (Jarod) Liang, a PhD student at the University of Technology Sydney, who were awarded travel grants to attend the XXVIIIth International Biometric Conference in Victoria, Canada. Australasian Applied Statistics Conference Join us for AASC’16 and pre-conference workshops, being held from 28 November - 2 December 2016 at The Windsong Pavilion, Four Winds, Barragga Bay, NSW, Australia.The themes of AASC 16 are “Big data and innovative consulting, statistics in fisheries and ecology, statistical genetics and modern approaches to smoothing complex data”. An exciting group of invited speakers will help explore this theme in various contexts. Bermagui is quoted as a “Destination that borders on perfection”. It is world renowned for its natural beauty and unspoiled beaches and forests. It offers a unique and exciting range of tourist attractions and is the gateway to the northern entrances to the Sapphire Coast of NSW. For more information visit: http://aasc.org.au/. Vanessa Cave British & Irish Region (BIR) 30-year Anniversary Meeting of DerSimonian & Laird On Wednesday, 13 April, a celebration was held to mark the 30th anniversary since the publication of the landmark paper “Metaanalysis in clinical trials” by DerSimonian & Laird (1986, Controlled Clinical Trials 7, 177-188). The paper is one of the most highly cited research publications. This paper introduces the now standard random-effects model for meta-analysis, a variety of estimation methods and successfully applies these to a variety of real datasets. The celebration brought together statistical scientists from across the world and included talks by Dr. Julian Higgins (University of Bristol), Dr. Kerry Dawn (University of Liverpool), Dr. Richard Riley (Keele University) and Dr. Theo Stijnen (Leiden University Medical Centre). It was a fantastic event for all involved, and was attended by both Nan Laird and Rebecca DerSimonian. Julian Higgins introduced the paper and described the history relating to it. He described the paper as “A beautifully written paper with a wealth of extremely sensible discussion of meta-analysis” but also emphasised that it was “not without its detractors!” Kerry discussed the impact of the paper on groups like the Cochrane Collaboration and also described the exciting future 11 Nan Laird showing a heterogeneous dataset from New Cuyama. Roland Langrock Dutch Region (ANed) Spring Meeting – Biostat’s Better Half, 24 June 2016, Leiden, the Netherlands The next annual spring meeting, Biostat’s Better Half, promises to be an entertaining event: in addition to the Dutch Region’s Annual Meeting, Biostatistics’ Better Half will be presenting her latest research. The keynote speakers will be Nicole Augustin (University of Bath), Nan van Geloven (Leiden UMC), Helene Jacqmin-Gadda (University of Bordeaux), Mar Rodriguez Girondo (Leiden UMC) and Sabine Schnabel (Wageningen University). During the meeting, also the Hans van Houwelingen Award for the best Dutch paper in a refereed journal in the biometrical field in 2014 and 2015 will be awarded. Registration is free. (https://forms.lumc.nl/lumc2/BMS) Joanna in ‘t Hout Eastern Mediterranean Region (EMR) Bulgaria We are pleased to announce that a short course on Survival Analysis will be held at the Faculty of Mathematics and Informatics, Sofia University, Sofia, Bulgaria, July 18 – 20, 2016. The speakers are Laura Antolini (Associate Professor in Medical Statistics, Department of Medicine and Surgery) and Davide Bernasconi (Post-doctoral fellow at the Department of Health Sciences) from the University of MilanoBicocca, Italy. The course is financially supported by the International Biometric Society and is open to everyone interested free of charge.There is no space limitation, but in order to attend the course Biometric Bulletin you need to register via the following registration form: http://goo.gl/ forms/AIsmZZXmQG1OKcms2. The deadline for registration is July 8, 2016. The tentative program of the short course is as follows: Day 1 – 07/18/2016 • Features of survival data • Random variables and theoretical functions • Nonparametric estimation of hazard and survival functions • Non parametric testing and sample size and power analysis • R tutorials Day 2 – 07/19/2016 • Competing risks data • Nonparametric estimation of hazard and crude incidence functions • Multistate data • Nonparametric estimation of hazard functions and state probabilities • R tutorials Day 3 – 07/20/2016 • Proportional hazard model: Cox model and fully parametric models • Check of assumptions • Stratified Cox model • Cox model with time varying effects • Cox model with time varying covariates • Fine and Gray model • R tutorials Practical computer classes in R are part of the course and some knowledge of R is preferable. If you have any inquiries (formal or informal), please do not hesitate to contact the main organizer Denitsa Grigorova ([email protected]), the Bulgarian representative of the Eastern Mediterranean Region of the International Biometric Society. Greece The EMR-IBS member and former Editor of the Biometric Bulletin, Professor Ranny Dafni, was appointed to the Board of Directors of the Hellenic Centre for Disease Control and Prevention (HCDCP) (http://www.keelpno.gr). HCDCP has been really active lately during the ongoing refugee crisis, overviewing vaccination programs and disease control at refugee camps. The 29th Annual Conference of the Greek Statistical Institute (http:// www.esi-stat.gr) has taken place in the town of Naoussa, Imathia prefecture, Greece from May 4 – 7, 2016. The conference focused on ‘Environmental Risk Assessment,’ but there were sessions covering a wide range of statistical theory and applications topics. About 100 participants took part and also enjoyed the beauties and tastes of this fabulous wine producing area of the Central Macedonia region. Israel The Israeli Statistical Association held its annual meeting in RamatEfal on May 19, 2016. There were about 200 participants from universities, the industry and governmental offices in Israel, including members of the Israeli Biostatistics Forum (IBF). The Program Committee included Dr. Yuval Nov (Chair) and Dr. Yair Goldberg (Haifa University), Dr. Havi Murad (Editor of Biometric Bulletin, Gertner Institute), and Prof. Sharon Rosset (Tel-Aviv University). The day opened with a keynote talk by Prof.Trevor Hastie (Stanford University) titled “Statistical Learning with Big Data”. There were a variety of sessions. For example, Bioinformatics and Survival Analysis (EMR-IBS sessions), which attracted members of the IBF. In many of the talks in these sessions a topic that was mentioned in different contexts was how to deal with the violation of two assumptions simultaneously: (i) random sampling (for example in case-control studies when you use ascertainment and sample more cases than in the population, and (ii) independent observations (i.e when you collect data on relatives from the same family). Other sessions were Statistical Inference and Learning, Statistics and Public Policy, Statistical Methods for the Analysis of Air Pollution Data, and a Poster Session (with prizes). The meeting concluded with a panel discussion on “Statistics and Data Science – Setting a New Course?”. The panel included Professor Trevor Hastie (Stanford University), Professor Avishai Mandelbaum (Technion), and Dr. Ronny Lempel (VP Recommendations Group at Outbrain, an online advertiser specializing in presenting sponsored website links). The panel was moderated by Dr. Itai Dattner (Haifa University). The panel members discussed the statisticians’ roles in the big data era, how we can become more dominant, and how it should affect our curriculum. Panel discussants from left to right: Ronny Lempel (Outbrain), Avishai Mandelbaum (Technion), Trevor Hastie (Stanford). A day after the conference, on May 20, Professor Hastie gave a four hour workshop on “Statistical Learning and Data Science”. Statistipedia – a competition on improving selected definitions in ‘Statistics and Probability’ from the Hebrew website of Wikipedia, initiated by the Israeli Statistical Association 20.10.15 – 20.12.15. The competition was open to students from all the faculties and also to high school students. Two prizes of $520 and $775 were awarded to the first two winners during the Israeli Statistical Association conference as well as 6 appreciation awards of $130 each. Winners of Statistipedia. 9th EMR-IBS Conference Thessaloniki The 9th Eastern Mediterranean Region – International Biometric Society Conference will be held in Thessaloniki, Greece from 8–12 Biometric Bulletin 12 May 2017. As always, we would like to bring together researchers from around the world to this beautiful venue. The program will include sessions from other regions of IBS, student travel awards, invited talks and some pre-conference courses, as in previous EMR meetings. Thessaloniki is considered an international city, a crossroad of different civilizations, with historical sites related to all countries of EMR. Thessaloniki is also a beautiful city with a combination of a lot of different activities and sightseeing. The entire meeting is devoted to the memory of Prof. Marvin Zelen (Harvard University), a keen supported of EMR, who passed away in November 2014. Symposium Honoring Prof. Marvin Zelen, Thessaloniki, Greece Frontier Science Foundation Hellas (FSFH), a non-profit organization, is organizing a symposium to honor Marvin Zelen, Co-founder of FSFH from 6–7 May 2017, adjacent to the 9th EMR-IBS Conference. Frontier Science & Technology Research Foundation, Inc. (FSTRF) and Frontier Science Scotland (FSS) are the Co-sponsors of the symposium. Details will be posted at: www.frontier-science.gr. Giota Touloumi Eastern North American Region (ENAR) WebENAR Be sure to check the ENAR Webinar website for updates regarding the upcoming WebENAR series, as well as for links to past WebENAR: http://www.enar.org/education/index.cfm. 2016 JSM 30 July-4 August, Chicago, IL, USA The 2016 JSM will convene at McCormick Place in Chicago, Illinois from 30 July–4 August. The theme of the 2016 meeting is “The Extraordinary Power of Statistics”. ENAR has been instrumental in helping to put together an outstanding program, including sponsoring many sessions. These sessions include invited presentations about analytics for personalized medicine, integrative genomics, advances in spatial and spatio-temporal data, cancer genomics, evaluating biomarkers for precision medicine, dynamic prediction in clinical survival analysis, dynamic treatment regimes, big and complex data inference, enrichment designs in clinical development, and challenges of analyzing high-dimensional and big data. ENAR has also co-sponsored many contributed sessions, as well as special presentations, such as: introductory overview lectures (adaptive clinical trial design, data science, casual inference, spatio-temporal data analysis), the ASA President’s Invited Address (“Science and News: A Marriage of Convenience”, delivered by Joe Palca from National Public Radio), the Deming Lecture (“Profound Knowledge from a Knowledge Use Perspective”, delivered by Vincent P. Baraba from Market Insight Corporation), the ASA President’s Address (“Appreciating Statistics”, delivered by ASA President Jessica Utts from UC Irvine), and the Fisher Lecture (“Personalizing Disease Prevention: Statistical Challenges”, delivered by Allice S. Whittemore from Stanford University School of Medicine). New this year is the JSM Data Art Show, exploring the ar t in data. ENAR received many proposals for invited and topic-contributed sessions, and thanks everyone who put for th an idea. ENAR extends a huge thank you to Bin Nan from the University of Michigan for serving on the Program Committee for the 2016 13 JSM. For more details about the upcoming JSM meetings, please see: https://www.amstat.org/meetings/jsm/2016/index.cfm. 2017 ENAR Spring Meeting, 12-15 March, Washington, DC, USA The 2017 ENAR Spring Meeting will take place in Washington, DC, to be held at the Washington Hilton.The Program Committee is soliciting suggestions for invited paper sessions. Please suggest ideas and potential speakers and/or develop a formal proposal. Proposals on topics that have broad potential scientific impact are particularly encouraged. Invited sessions are 105 minutes long, and different formats are encouraged, such as sessions with four speakers, three speakers plus a discussant, or a panel discussion. The invited session proposals will be selected by the Program Committee, which includes the Program Chair, Associate Chair, 15 ASA representatives of ASA sections, and two ENAR at-large members. One participant may be a speaker/ panelist in at most one invited or contributed session. The deadline for submission is 15 June 2016. Formal invited session proposals can be submitted online through ENAR’s website. To informally suggest ideas, topics or names of potential speakers, contact Nandita Mitra ([email protected]), Program Chair or Andrea Foulkes ([email protected]), Associate Chair. 2017 JSM 29 July-3 August, Baltimore, MD, USA The 2017 Joint Statistical Meetings will be held in Baltimore, Maryland, and ENAR is fortunate to have Dionne Price be our representative to the Program Committee. If you have ideas for the meeting, feel free to contact Bin: [email protected]. 2018 ENAR Spring Meeting, 25-28 March, Atlanta, GA, USA Stay tuned for information about the 2018 ENAR Spring Meeting in Atlanta, Georgia! Leslie McClure German Region (DR) DAGStat 2016 The 4th Joint Meeting of the “Deutschen Arbeitsgemeinschaft Statistik” took place from March 14 to 18, 2016 at the GeorgAugust-University of Göttingen and included the Biometric Colloquium of the German Region (IBS-DR), the “Pfingsttagung” of the “Deutschen Statistischen Gesellschaft” (DStatG) and the Annual Meeting of the “Gesellschaft für Klassifikation” (GfKl). A total of 709 statisticians from 25 countries came to Göttingen, and again an increase in the number of participants was recorded compared to the previous three meetings in Freiburg, Dortmund and Bielefeld. The local organization was coordinated by the Chairs of Statistics and Econometrics, together with the Centre for Statistics of the University of Göttingen under the direction of Thomas Kneib and Henriette Irmer. From 451 submissions, a total of 368 oral presentations were selected that could be assigned to 33 topics. Psychometrics, Machine Learning, Bioinformatics and Big Data and Data Science were included for the first time into the program. The topics ‘Statistics in Clinical and Preclinical Research’ and ‘Survival and Event History Analysis’ were the most favorable, each with eight sessions. Most of the 30 invited speakers were from outside Germany and they contributed Biometric Bulletin to the international and scientific quality of the program, as did the over 100 international participants. The Statistics Bazaar, firstly introduced to the DAGStat meetings in Freiburg, was again an opportunity for early birds to learn about methods in an unfamiliar context. Among the 175 young scientists, especially the beginners were supported in the Young Statisticians’ Session that was jointly organized by the three Societies IBS-DR, DStatG and GfKl. A marzipan competition (organized by Udo Rempe) promoted the communication between the presenters by rating each other. Further scientific exchange was encouraged at the Welcome Reception in the historic rooms of the old refectory. Under the quote from Novalis, Freiherr von Hardenberg “Even the coincidence is not unfathomable, it has its regularity” the conference dinner was held in the Castle Hotel in Nörten-Hardenberg. The awarding of the DAGStat medals in a festive setting was embedded in lively scientific and non-scientific exchange between the participants.The social program was completed by guided city tours, with special emphasis on Carl Friedrich Gauss. In order to reach the public and provide information on statistical topics, the meeting started off with a public lecture in the historic building of the University Library, the Paulinerkirche. Ralf Wagner of the University of Kassel spoke on the theme, “Yes, Made for Me” or “Made from My Data!” - Targeting Customers in Interactive Marketing. The IBS-DR chose two young scientists to present their master thesis Young Statisticians’ Session: Anke Hüls (Chair), Moritz Hanke (IBS-DR), Maria Umlauft (GfKL), Martin Happ (DstatG), Tanja Proctor (IBS-DR), Benjamin Hofner (Chair) (from left to right). and reimbursed their conference participation and the conference dinner. In addition, IBS-DR awarded Susanne Steinhauser with the Bernd Streitberg Prize and Stefan Englert with the Gustav Adolf Lienert Prize. Both got the opportunity to present together with Andre Burkovski (Best Paper Award from the GfKl) in the Awards Session. Overall, the DAGStat Conference 2016 hopefully remains as a good memory, so that we will again meet in 2019 in München. We thank all supporters of the meeting and in particular the city of Göttingen for the financial support of the Welcome Reception and the Deutsche Forschungsgemeinschaft (DFG) and the program Pro * Niedersachsen funded by Lower Saxony’s Ministry of Science and Culture for the promotion of the scientific program of the meeting. (Henriette Irmer, Thomas Kneib) Axel Benner Indian Region (IR) Indiaclen Annual Conference SRM University, Chennai organized the 2016 Annual Indiaclen Conference March 11 – 12. Dr. K R John, Professor of Community Medicine was the Organizing Secretary. The conference was inaugurated by Dr. N K Arora, Global Executive Director of Inclen Network Trust. During the statistical session, Dr. Anil C Mathew, Professor of Biostatistics spoke on “Cox proportional hazards model, its characteristics and its application in clinical research”, and Dr. Thambu David, Professor CMC Vellore spoke on “Evidence based medicine”. The aim of the conference was to promote clinical epidemiology and improve the health in India. Awards Session: Berthold Lausen (GfKL), Andre Burkovski (GfKL best paper award), Susanne Steinhauser (Bernd Streitberg Price), Stefan Englert (Gustav Adolf Lienert Price), Tim Friede (IBS-DR) (from left to right). The IBS-DR lecture series “Education for Statistics in Practice” was held this year by Georg Heinze and Daniela Dunkler from the Medical University Wien on Variable Selection for Statistical Models. As in the previous years, the lecture has been a great success. Due to the excellent experience, the series will be extended to the International Biometric Conference. At the panel discussion, the Future of Scientific Publishing was discussed, moderated by the Chair of the panel, Karl Mosler. The 72 posters, combined with a reception on the occasion of the 40th Annual Meeting of the GFKl, promoted scientific exchange in an informal setting. This time, two rating systems rewarded the work of the poster presenters. A consortium determined five poster authors who obtained book vouchers generously donated by Springer Verlag. Dr. Anil C Mathew delivering the invited talk Cox proportional hazards model, its characteristics and its application in clinical research. Biometric Bulletin 14 34th Annual National Conference Indian Society for Medical Statistics (ISMS) The 34th Annual National Conference Indian Society for Medical Statistics is scheduled to be held from 1 – 3 December 2016 at the Indian Statistical Institute, Kolkata. For further details, you may contact Prof. Saurabh Ghosh by emailing [email protected]. Anil C Mathew Italian Region (RItl) The Italian Region is strongly involved in promoting the education in biostatistics to non-specialists, and for the next month we are promoting two short courses: 1. Statistics Applied to Biomedical Research, organized by the Italian Association of Medical Physics, 16 – 17 June 2016 at the University Milano Bicocca and BiostAT2016. 2. XXIII Summer School of Statistical Inference in Biology and Human Science, 28 June – 8 July 2016 at UNI-Astiss. We are also working on our homepage, www.ibs-italy.org, which will be set online 20 May 2016. Livio Finos Japanese Region (JR) The 2016 Annual Meeting of the Biometric Society of Japan The 2016 Annual Meeting of the Biometric Society of Japan was held on 18-19 March, 2016 at Institute of Statistical Mathematics, Tachikawa, Tokyo, Japan, with 196 participants. The Biometric Society of Japan is discussing establishing a certificate system for biostatisticians working in clinical trials. A special session on the certificate system was held. The working group for the certification system reported the current progress, and several issues were discussed. In an invited session, recent topics in statistical methods for pharmaceutical research were discussed. The Biometric Society of Japan conferred the Young Biostatisticians Award on Drs. Fumiaki Kobayashi (Daiichi Sankyo, Inc.), Osamu Komori (Fukui University) and Akihiro Hirakawa (Nagoya University) for their publications in Journal of Agricultural, Biological and Environmental Statistics, Biometrics and Japanese Journal of Biometrics, respectively. A tutorial session entitled “Introduction to statistical modeling: from the generalized linear model to the Bayesian hierarchical models” was held with 106 participants. The 2016 Japanese Joint Statistical Meeting The winners of the Young Biostatisticians Award: Drs. Koyama (for Dr. Kobayashi), Hirakawa and Komori (from left to right). The Biometric Society of Japan is one of the six sponsoring organizations of the meeting, and the 2016 Japanese Joint Statistical Meeting will be held on 4-7 September at Kanazawa University in Kanazawa, Japan. The Biometric Society of Japan is organizing the Biometric Symposium entitled “Statistical monitoring for quality control and assurance in clinical trials”. Recently, inappropriate handling of data in medical researches has been a serious social issue in Japan. The symposium will cover statistical methods for safety monitoring and fraud detection, which is expected to be useful to detect distortion in data handling and analysis. The Biometric Society of Japan is also organizing a session by the winners of the Young Biostatisticians Award conferred by the Society. Three young statisticians will make a presentation on their research. Satoshi Hattori Western North American Region (WNAR) WNAR is the local host to the 2016 IBC, to be held in Victoria, Canada from 10-15 July 2016. Meeting information is available at: http://biometricconference.org. Details about WNAR sessions at IBC and the WNAR student competitions will be posted on the WNAR web page www.wnar.org as they become available. Megan Othus IBS on LinkedIn - Join Our Group & Get Connected to Colleagues Across the Globe IBS has created a LinkedIn Group for biometrics industry professionals to become a part of…and network with your colleagues instantaneously! Post discussions to the Group and get comments/feedback from Group members on their perspectives or experiences. A great benefit of this Group is that it’s a very easy and free way to communicate with 15 your colleagues who live all around the world. Being connected to the IBS Group will allow you to see other connection possibilities as well and broaden your professional network. The possibilities are endless. Join our Group today by visiting www.linkedin.com, and search under Groups for ‘International Biometric Society.’ Biometric Bulletin MEETINGS 2016 24 June 28 November – 2 December 29 June – 1 July 1 – 3 December Australasian Applied Statistics Conference (AASC’16) Barragga Bay, NSW, Australia http://aasc.org.au/ Spring Meeting – Biostat’s Better Half Leiden, the Netherlands Registration is free! (https://forms.lumc.nl/lumc2/BMS) Summer School on Group Sequential and Adaptive Study Designs Strobl at the lake “Wolfgangssee”, Austria 34th Annual National Conference Indian Society for Medical Statistics (ISMS) Indian Statistical Institute, Kolkata, India Prof. Saurabh Ghosh ([email protected]) 28 June – 8 July XXIII Summer School of Statistical Inference in Biology and Human Science UNI-Astiss, Italy 2017 10 – 15 July 12 – 15 March XXVIIIth International Biometric Conference Victoria, BC, Canada http://biometricconference.org/ ENAR Spring Meeting Washington, DC, USA http://www.enar.org/meetings/index.cfm 18 – 20 July 8 – 12 May 9th EMR-IBS Conference Thessaloniki, Greece Short course on Survival Analysis Sofia, Bulgaria (sponsored by EMR-IBS) Registration is free! (http://goo.gl/forms/ AIsmZZXmQG1OKcms2) Denitsa Grigorova ([email protected]) 29 July – 3 August Joint Statistical Modeling Baltimore, MD, USA Dionne Price ([email protected]) July 30 – August 4 Joint Statistical Meeting Chicago, IL, USA http://www.enar.org/meetings/index.cfm August 21 – 25 International Society for Clinical Biostatistics 38th Annual Conference Vigo, Spain http://www.iscb.info/ 21 – 25 August International Society for Clinical Biostatistics 37th Annual Conference Birmingham, United Kingdom http://www.iscb.info/ 5 – 6 September 2018 2nd Workshop for Young Statisticians of the Spanish Biometric Society Barcelona, Spain http://www.biometricsociety.net/ii-jornadas-cientificas-de-estudiantes-de-la-seb/ 4 – 7 September 25 – 28 March ENAR Spring Meeting Atlanta, GA, USA 8 – 13 July XXIXth International Biometric Conference Barcelona, Spain The 2016 Japanese Joint Statistical Meeting Kanazawa University, Kanazawa, Japan Biometric Bulletin 16