Polyfit-package: Polyfit add-on to DESeq
In Polyfit: Add-on to DESeq to improve p-values and q-values

Description Details Author(s) References Examples

implementation the Polyfit add-on to DESeq described in the paper "Improved error estimates for the analysis of differential expression from RNA-seq data"

Package:	Polyfit
Type:	Package
Version:	0.99.3
Date:	2014-08-06
License:	GPL(>=3)

Polyfit is an add-on to the negative-binomial based packages DESeq for two-class detection of differential expression which ensures the p-value distribution is uniform over the interval [0, 1] for data satisfying the null hypothesis of no differential expression. The first component is the function pfNbinomTest which replaces the function nbinomTest in DESeq. Its purpose is to smooth point singularities, particularly one at p = 1, in the p-value distribution caused by calculating calculating p-values from a discrete distribution. The output from this function should then be passed to the second component, the function link{levelPValues}. Its purpose is to apply a variant of the Storey-Tibshirani procedure to shift the p-values so that those corresponding to the null hypothesis have a unform distribution, and to calculate corresponding q-values (or 'adjusted p-values') for controlling errors via the false discovery rate.

Conrad Burden

Maintainer: conrad.burden@anu.edu.au

Burden, C.J., Qureshi, S. and Wilson, S.R. (2014). Error estimates for the analysis of differential expression from RNA-seq count data, PeerJ PrePrints 2:e400v1.

Robinson, M., McCarthy, D., and Smyth, G. (2010). edgeR: a Bioconductor package for differential expression analysis of digital gene expression data. Bioinformatics, 26, 139-140.

Anders, S. and Huber, W. (2010). Differential expression analysis for sequence count data. Genome Biology, 11(10), R106.

# Example using DESeq
cds <- makeExampleCountDataSet()
cds <- estimateSizeFactors( cds )
cds <- estimateDispersions( cds )
nbTPolyfit <- pfNbinomTest( cds, "A", "B" )

lP <- levelPValues(nbTPolyfit$pval)
pvalTab <- cbind(origPval=nbTPolyfit$pval, correctedPval=lP$pValueCorr, qval=lP$qValueCorr)
cat("\n Original and corrected P-values from DESeq \n")
head(pvalTab)