Jon Wakefield1. 1. Departments of Statistics and Biostatistics, University of Washington, Seattle, USA. jonno@u.washington.edu
Abstract
BACKGROUND: In the context of genome-wide association studies we critique a number of methods that have been suggested for flagging associations for further investigation. METHODS: The P-value is by far the most commonly used measure, but requires careful calibration when the a priori probability of an association is small, and discards information by not considering the power associated with each test. The q-value is a frequentist method by which the false discovery rate (FDR) may be controlled. RESULTS: We advocate the use of the Bayes factor as a summary of the information in the data with respect to the comparison of the null and alternative hypotheses, and describe a recently-proposed approach to the calculation of the Bayes factor that is easily implemented. The combination of data across studies is straightforward using the Bayes factor approach, as are power calculations. CONCLUSIONS: The Bayes factor and the q-value provide complementary information and when used in addition to the P-value may be used to reduce the number of reported findings that are subsequently not reproduced.
BACKGROUND: In the context of genome-wide association studies we critique a number of methods that have been suggested for flagging associations for further investigation. METHODS: The P-value is by far the most commonly used measure, but requires careful calibration when the a priori probability of an association is small, and discards information by not considering the power associated with each test. The q-value is a frequentist method by which the false discovery rate (FDR) may be controlled. RESULTS: We advocate the use of the Bayes factor as a summary of the information in the data with respect to the comparison of the null and alternative hypotheses, and describe a recently-proposed approach to the calculation of the Bayes factor that is easily implemented. The combination of data across studies is straightforward using the Bayes factor approach, as are power calculations. CONCLUSIONS: The Bayes factor and the q-value provide complementary information and when used in addition to the P-value may be used to reduce the number of reported findings that are subsequently not reproduced.
Authors: Darren R Brenner; Christopher I Amos; Yonathan Brhane; Maria N Timofeeva; Neil Caporaso; Yufei Wang; David C Christiani; Heike Bickeböller; Ping Yang; Demetrius Albanes; Victoria L Stevens; Susan Gapstur; James McKay; Paolo Boffetta; David Zaridze; Neonilia Szeszenia-Dabrowska; Jolanta Lissowska; Peter Rudnai; Eleonora Fabianova; Dana Mates; Vladimir Bencko; Lenka Foretova; Vladimir Janout; Hans E Krokan; Frank Skorpen; Maiken E Gabrielsen; Lars Vatten; Inger Njølstad; Chu Chen; Gary Goodman; Mark Lathrop; Tõnu Vooder; Kristjan Välk; Mari Nelis; Andres Metspalu; Peter Broderick; Timothy Eisen; Xifeng Wu; Di Zhang; Wei Chen; Margaret R Spitz; Yongyue Wei; Li Su; Dong Xie; Jun She; Keitaro Matsuo; Fumihiko Matsuda; Hidemi Ito; Angela Risch; Joachim Heinrich; Albert Rosenberger; Thomas Muley; Hendrik Dienemann; John K Field; Olaide Raji; Ying Chen; John Gosney; Triantafillos Liloglou; Michael P A Davies; Michael Marcus; John McLaughlin; Irene Orlow; Younghun Han; Yafang Li; Xuchen Zong; Mattias Johansson; Geoffrey Liu; Shelley S Tworoger; Loic Le Marchand; Brian E Henderson; Lynne R Wilkens; Juncheng Dai; Hongbing Shen; Richard S Houlston; Maria T Landi; Paul Brennan; Rayjean J Hung Journal: Carcinogenesis Date: 2015-09-10 Impact factor: 4.944
Authors: George R Uhl; Tomas Drgon; Catherine Johnson; Chuan-Yun Li; Carlo Contoreggi; Judith Hess; Daniel Naiman; Qing-Rong Liu Journal: Ann N Y Acad Sci Date: 2008-10 Impact factor: 5.691
Authors: Katsuhito Yasuno; Kaya Bilguvar; Philippe Bijlenga; Siew-Kee Low; Boris Krischek; Georg Auburger; Matthias Simon; Dietmar Krex; Zulfikar Arlier; Nikhil Nayak; Ynte M Ruigrok; Mika Niemelä; Atsushi Tajima; Mikael von und zu Fraunberg; Tamás Dóczi; Florentina Wirjatijasa; Akira Hata; Jordi Blasco; Agi Oszvald; Hidetoshi Kasuya; Gulam Zilani; Beate Schoch; Pankaj Singh; Carsten Stüer; Roelof Risselada; Jürgen Beck; Teresa Sola; Filomena Ricciardi; Arpo Aromaa; Thomas Illig; Stefan Schreiber; Cornelia M van Duijn; Leonard H van den Berg; Claire Perret; Carole Proust; Constantin Roder; Ali K Ozturk; Emília Gaál; Daniela Berg; Christof Geisen; Christoph M Friedrich; Paul Summers; Alejandro F Frangi; Matthew W State; H Erich Wichmann; Monique M B Breteler; Cisca Wijmenga; Shrikant Mane; Leena Peltonen; Vivas Elio; Miriam C J M Sturkenboom; Patricia Lawford; James Byrne; Juan Macho; Erol I Sandalcioglu; Bernhard Meyer; Andreas Raabe; Helmuth Steinmetz; Daniel Rüfenacht; Juha E Jääskeläinen; Juha Hernesniemi; Gabriel J E Rinkel; Hitoshi Zembutsu; Ituro Inoue; Aarno Palotie; François Cambien; Yusuke Nakamura; Richard P Lifton; Murat Günel Journal: Nat Genet Date: 2010-04-04 Impact factor: 38.330