Empirical bayes methods and false discovery rates for microarrays

By Bradley Efron, Robert Tibshirani. Genetic Epidemiology, 2002

725 citations recorded by the source corpus.

Subfields: Molecular Biology, Statistics and Probability

View Paper: https://www.semanticscholar.org/paper/d2edf7654e05615b5d769c5e969a4b9030605f1b

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Molecular Biology

1. Bradley Efron
2. Gordon K. Smyth
3. A. Chakrabarti
4. Jayanta Kumar Ghosh
5. Hamish S. Scott
6. Joëlle Michaud
7. David Broadhurst
8. Masanao Yajima

In Statistics and Probability

1. Andrew Gelman
2. A. Chakrabarti
3. Jennifer L. Hill
4. John P. A. Ioannidis
5. Douglas B. Kell
6. David Broadhurst
7. Jayanta Kumar Ghosh
8. Masanao Yajima

Citations in the corpus

• Estimation of Significance Thresholds for Genomewide Association Scans (2008, background): “More recently, the false discovery rate (FDR) [Benjamini and Hochberg, 1995] and variations [Efron and Tibshirani, 2002] have gained support, as we may tolerate some type1 errors so long as they are a small proportion of the rejected hypotheses.”
• Use of within-array replicate spots for assessing differential expression in microarray experiments (2005, background): “…each individual gene can be made more reliable by making use of information generated from the whole ensemble of genes (Newton et al., 2001; Tusher et al., 2001; Efron et al., 2001; Efron and Tibshirani, 2002; L¨onnstedt and Speed, 2002; Kendziorski et al., 2003; Newton et al., 2004; Smyth, 2004).”
• Diverse RNA-Binding Proteins Interact with Functionally Related Sets of RNAs, Suggesting an Extensive Regulatory System (2008, methodology): “For two proteins in the survey (Ssd1 and Khd1), we used a more stringent 1% local FDR criterion [45] (details in Materials and Methods; Datasets S2 and S3).”
• An exploratory test for an excess of significant findings (2007, background): “The probability of finding a ‘positive’ result in a study is equal to the probability of a finding a true positive result plus the probability of finding a false positive result [5,36,41,42]:”
• Detecting subnetwork-level dynamic correlations (2017, methodology): “By combining LA with the egonetwork approach, and using mixture models for LA scores through local false discovery rate (lfdr) inference (Efron and Tibshirani, 2002), our method finds network regions where the LA scouting genes to a given ego-network are concentrated.”
• The Bayes oracle and asymptotic optimality of multiple testing procedures under sparsity (2010, methodology): “For tests with a fixed threshold, Efron and Tibshirani [14] define another very similar measure, called the Bayesian False Discovery Rate, BFDR: where t 1 and t 2 are the probabilities of type I and type II errors.”
• Calibration of computational tools for missense variant pathogenicity classification and ClinGen recommendations for PP3/BP4 criteria (2022, background): “36,37 Computational objective The objective of our approach was to develop a framework that discretizes the prediction range for any given computational tool into a set of nine intervals, corresponding to the various levels of evidential strength: supporting, moderate, strong, and very strong for…”
• Statistical strategies for avoiding false discoveries in metabolomics and related experiments (2007, background): “…of course, a characteristic of omics studies that has long been highlighted following the development of microarrays for transcriptomics (e.g. (Benjamini and Hochberg, 1995; Efron and Tibshirani, 2002; Storey, 2002; Storey and Tibshirani , 2003; Reiner et al., 2003; Jung, 2005; Xie et al., 2005)).”
• Identification of diagnostic subnetwork markers for cancer in human protein-protein interaction network (2010, background): “Extensive work has been done to identify differentially expressed genes across different phenotypes [1-5], which can be used as diagnostic markers for classifying different disease states or predicting the clinical outcomes [6-11].”
• Large-Scale Simultaneous Hypothesis Testing (2004, methodology): “Figure 6 relates to a well-known microarray experiment concerning differences between two types of genetic mutations causing increased breast cancer risk, “BRCA1” and “BRCA2”; see Hedenfalk et al. (2001), also Efron and Tibshirani (2002), and Efron (2003).”
• Escaping the Bonferroni iron claw in ecological studies (2004, background): “Research on the ‘‘FDR family’’ is currently very active, and new concepts, which apparently may improve the original idea, are currently emerging (Efron and Tibshirani 2002, Genovese and Wasserman 2002, Sarkar 2002, Fernando et al. 2004).”
• A comparison of the Benjamini-Hochberg procedure with some Bayesian rules for multiple testing (2008, background): “[8], Efron and Tibshirani [7], Storey et al.”
• Size, power and false discovery rates (2007, background): “…z ) corresponds to Storey’s “ q -value,” the tail-area false discovery rate attained at a given observed value z i = z . d The estimated fdr curve in Figure 2 is c fdr( z ) = p 0 f 0 ( z ) / b f ( z ) , (2.13) where f 0 ( z ) is the standard normal density ϕ ( z ) = exp z 2 / 2 / √ 2 x , p 0 = 0 .”
• Correlation and Large-Scale Simultaneous Significance Testing (2007, background): “Fdr( x ) is an empirical Bayes estimate of the a posteriori probability that case i is null given z i ≥ x , Storey (2002), Efron and Tibshirani (2002), amounting to a version of Storey’s “ q -value”.”
• Metabolomics and Multi-Omics Integration: A Survey of Computational Methods and Resources (2020, methodology): “Other methods tailored to multiple comparison corrections in omics data include the tail statistic, which is based on an expected distribution of p-values [141,142], and an Empirical Bayes method originally developed for microarrays but applicable to other omic data [143].”