Dataset Information

A general framework for multiple testing dependence.

ABSTRACT: We develop a general framework for performing large-scale significance testing in the presence of arbitrarily strong dependence. We derive a low-dimensional set of random vectors, called a dependence kernel, that fully captures the dependence structure in an observed high-dimensional dataset. This result shows a surprising reversal of the "curse of dimensionality" in the high-dimensional hypothesis testing setting. We show theoretically that conditioning on a dependence kernel is sufficient to render statistical tests independent regardless of the level of dependence in the observed data. This framework for multiple testing dependence has implications in a variety of common multiple testing problems, such as in gene expression studies, brain imaging, and spatial epidemiology.

SUBMITTER: Leek JT

PROVIDER: S-EPMC2586646 | biostudies-literature | 2008 Dec

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

A general framework for multiple testing dependence.

Leek Jeffrey T JT Storey John D JD

Proceedings of the National Academy of Sciences of the United States of America 20081124 48

We develop a general framework for performing large-scale significance testing in the presence of arbitrarily strong dependence. We derive a low-dimensional set of random vectors, called a dependence kernel, that fully captures the dependence structure in an observed high-dimensional dataset. This result shows a surprising reversal of the "curse of dimensionality" in the high-dimensional hypothesis testing setting. We show theoretically that conditioning on a dependence kernel is sufficient to r ...[more]

PMID: 19033188

Similar Datasets

Project description:Background: Peanut introduction guidelines have undergone significant reversal since 2001 from recommending delayed introduction to rescinding the recommendations in 2008 to actively recommending early introduction of peanut between 4 and 11 months of age in high-risk infants in 2015. This qualitative study aims to explore pediatrician beliefs, practices, facilitators, and barriers regarding peanut introduction and testing. Methods: General pediatricians from academic, private, large group, and underserved practices in Northern California underwent individual semi-structured interviews in 2017. We asked about experiences surrounding infant peanut introduction, strategies for staying up-to-date with current recommendations, and barriers and facilitators to the new peanut introduction and testing recommendations. The data were coded, and using grounded theory methodology, a conceptual framework was developed around early peanut introduction and testing in infants. Results: Eighteen general pediatricians participated. We identified barriers that may contribute to pediatrician reluctance to recommending early peanut introduction or testing including lack of awareness, lack of agreement, lack of resources, and lack of outcome expectancy. A framework was created that suggests that pediatricians need to be knowledgeable about new recommendations, agree with the recommendations, have resources to carry out the counseling and testing, and have buy-in from the parents in order for successful uptake of peanut introduction guidelines. Conclusion: Recommending early peanut introduction or testing causes significant apprehension in some pediatricians, and there are many barriers to following recent early peanut introduction recommendations. A potential limitation of the study is that it was conducted right after the addendum guidelines were changed, leaving the possibility that attitudes and practices may have evolved since 2017. It is still likely that a multifaceted approach that addresses primary care provider guideline awareness, limited primary care resources for education and testing, and includes support and collaboration from subspecialty practices is more likely to lead to improved early peanut introduction uptake.

Dataset Information

A general framework for multiple testing dependence.

Publications

A general framework for multiple testing dependence.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets