BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:A powerful empirical Bayes approach for high dimensional replicabi
 lity analysis  - Hongyuan Cao (Florida State University)
DTSTART:20230512T130000Z
DTEND:20230512T140000Z
UID:TALK199486@talks.cam.ac.uk
CONTACT:Qingyuan Zhao
DESCRIPTION:Researchers are interested in combining information across mul
 tiple (heterogeneous) studies to discover if findings are reproducible in 
 different populations or in different studies.  The goal is to increase th
 e power and control the false discovery rate\, which results in reliable a
 nd robust scientific findings.  The focus of our study is to draw inferenc
 es regarding a large number of features\, such as gene expression\, DNA me
 thylation\, and others.  Rather than combining the underlying raw data\, w
 hich is not always easy due to differences in the experimental designs\, m
 ost approaches\, including our proposed approach\, are based on p-values d
 erived from individual studies.  The popular approaches currently used in 
 the literature either cannot control the false discovery rate or have low 
 power since the null hypothesis of replicability analysis is a composite n
 ull hypothesis. We develop an empirical Bayes approach for the mixture mod
 el by jointly modeling the hidden states corresponding to null and alterna
 tive hypotheses across the studies. The method uses a non-parametric EM al
 gorithm combined with the pool-adjacent-violator-algorithm (PAVA).  In doi
 ng so\, our method borrows information across features and different studi
 es while accounting for heterogeneity. We demonstrate theoretically that t
 he proposed method controls the false discovery rate (FDR). Extensive simu
 lation studies show that the proposed method has higher power than the exi
 sting methods while controlling the FDR. Datasets from spatial transcripto
 mic studies are used to illustrate our methodology.  
LOCATION:MR12\, Centre for Mathematical Sciences
END:VEVENT
END:VCALENDAR