http://precedings.nature.com/tags/P-value Recently posted documents tagged with 'P-value' Nature Publishing Group en Nature Precedings http://precedings.nature.com/images/header_logo.gif http://precedings.nature.com http://precedings.nature.com/documents/2002/version/3 Most of the statistical tests currently used to detect differentially expressed genes are based on asymptotic results, and perform poorly for low expression tags. Another problem is the common use of a single canonical cutoff for the significance level (p-value) of all the tags, without taking into consideration the type II error and the highly variable character of the sample size of the tags.This work reports the development of two significance tests for the comparison of digital expression profiles, based on frequentist and Bayesian points of view, respectively. Both tests are exact, and do not use any asymptotic considerations, thus producing more correct results for low frequency tags than the chi-square test. The frequentist test uses a tag-customized critical level which minimizes a linear combination of type I and type II errors. A comparison of the Bayesian and the frequentist tests revealed that they are linked by a Beta distribution function. These tests can be used alone or in conjunction, and represent an improvement over the currently available methods for comparing digital profiles. http://precedings.nature.com/documents/2002/version/3 Fri, 29 Aug 2008 21:08:06 UTC Significance tests for comparing digital gene expression profiles hdl:10101/npre.2008.2002.3 2009-09-03 Leonardo Varuzza Nature Precedings 2008-08-29T21:08:06Z Manuscript Genetics & Genomics Bioinformatics http://creativecommons.org/licenses/by/3.0/ http://dx.doi.org/10.1038/npre.2008.2002.2 Most of the statistical tests currently used to detect differentially expressed genes are based on asymptotic results, and perform poorly for low expression tags. Another problem is the common use of a single canonical cutoff for the significance level (p-value) of all the tags, without taking into consideration the type II error and the highly variable character of the sample size of the tags.This work reports the development of two significance tests for the comparison of digital expression profiles, based on frequentist and Bayesian points of view, respectively. Both tests are exact, and do not use any asymptotic considerations, thus producing more correct results for low frequency tags than the chi-square test. The frequentist test uses a tag-customized critical level which minimizes a linear combination of type I and type II errors. A comparison of the Bayesian and the frequentist tests revealed that they are linked by a Beta distribution function. These tests can be used alone or in conjunction, and represent an improvement over the currently available methods for comparing digital profiles. http://dx.doi.org/10.1038/npre.2008.2002.2 Wed, 27 Aug 2008 15:01:43 UTC Significance tests for comparing digital gene expression profiles doi:10.1038/npre.2008.2002.2 2008-08-27 Leonardo Varuzza Nature Precedings 2008-08-27T15:01:43Z Manuscript Developmental Biology Bioinformatics http://creativecommons.org/licenses/by/3.0/ http://dx.doi.org/10.1038/npre.2008.2002.1 This paper is about differential gene expression measured by transcript counting methods such as SAGE or MPSS. It introduces two significance tests for detection of differential expressed tags: frequentist and Bayesian. Under the frequentist view, it is proposed a test that computes the critical level as a function of each tag total frequency. Under the Bayesian view the Full Bayesian Significance Test is used considering the logistic normal distribution. The two proposed significance levels, the frequentist and the Bayesian, are compared for a data set with four libraries. The linking function between them is a Beta distribution function with mean 0.39 and standard deviation 0.30. http://dx.doi.org/10.1038/npre.2008.2002.1 Mon, 23 Jun 2008 14:23:29 UTC Comparative Enumeration Gene Expression doi:10.1038/npre.2008.2002.1 2008-06-23 Leonardo Varuzza Nature Precedings 2008-06-23T14:23:29Z Manuscript Developmental Biology Bioinformatics http://creativecommons.org/licenses/by/3.0/ http://precedings.nature.com/documents/306/version/2 Reproducibility is a fundamental requirement in scientific experiments and clinical contexts. Recent publications raise concerns about the reliability of microarray technology because of the apparent lack of agreement between lists of differentially expressed genes (DEGs). In this study we demonstrate that (1) such discordance may stem from ranking and selecting DEGs solely by statistical significance (P) derived from widely used simple t-tests; (2) when fold change (FC) is used as the ranking criterion, the lists become much more reproducible, especially when fewer genes are selected; and (3) the instability of short DEG lists based on P cutoffs is an expected mathematical consequence of the high variability of the t-values. We recommend the use of FC ranking plus a non-stringent P cutoff as a baseline practice in order to generate more reproducible DEG lists. The FC criterion enhances reproducibility while the P criterion balances sensitivity and specificity. http://precedings.nature.com/documents/306/version/2 Tue, 03 Jul 2007 12:18:21 UTC The Reproducibility of Lists of Differentially Expressed Genes in Microarray Studies hdl:10101/npre.2007.306.2 2007-07-03 Leming D. Shi Nature Precedings 2007-07-03T12:18:21Z Manuscript Biotechnology Bioinformatics http://creativecommons.org/licenses/by/2.5/ http://dx.doi.org/10.1038/npre.2007.306.1 Reproducibility is a fundamental requirement in scientific experiments and clinical contexts. Recent publications raise concerns about the reliability of microarray technology because of the apparent lack of agreement between lists of differentially expressed genes (DEGs). In this study we demonstrate that (1) such discordance may stem from ranking and selecting DEGs solely by statistical significance (P) derived from widely used simple t-tests; (2) when fold change (FC) is used as the ranking criterion, the lists become much more reproducible, especially when fewer genes are selected; and (3) the instability of short DEG lists based on P cutoffs is an expected mathematical consequence of the high variability of the t-values. We recommend the use of FC ranking plus a non-stringent P cutoff as a baseline practice in order to generate more reproducible DEG lists. The FC criterion enhances reproducibility while the P criterion balances sensitivity and specificity. http://dx.doi.org/10.1038/npre.2007.306.1 Mon, 02 Jul 2007 09:20:53 UTC The Reproducibility of Lists of Differentially Expressed Genes in Microarray Studies doi:10.1038/npre.2007.306.1 2007-07-02 Leming Shi Nature Precedings 2007-07-02T09:20:53Z Manuscript Biotechnology Bioinformatics http://creativecommons.org/licenses/by/2.5/