tweeDE: Score test for differences between two Poisson-Tweedie groups

In tweeDEseq: RNA-seq data analysis using the Poisson-Tweedie family of distributions

Description

Carry out a score test for differences between two Poisson-Tweedie groups.

Usage

tweeDE(object, group, mc.cores = 1, pair = NULL, a = NULL, ...)
testPoissonTweedie(x, group, saveModel = FALSE, a = NULL, log = FALSE,  ...)
MAplot(x, ...)
Vplot(x, ...)
## S3 method for class 'tweeDE'
print(x, n=6L, sort.by="pval", log2fc.cutoff=0, pval.adjust.cutoff=1, print=TRUE, ...)
## S3 method for class 'tweeDE'
MAplot(x, log2fc.cutoff=0, highlight=NULL, ...)
## S3 method for class 'tweeDE'
Vplot(x, log2fc.cutoff=0, pval.adjust.cutoff=1, highlight=NULL, ylab=expression(paste(-log[10], " Raw P-value")), ...)

Arguments

object	a data.frame or a matrix of RNA-seq counts.
group	vector giving the experimental group/condition for each sample/library.
mc.cores	number of cpu cores to be used. This option is only available when the 'multicore' package is installed and loaded first. In such a case, if the default value of mc.cores=1 is not changed, all available cores will be used.
pair	vector of two elements containing the representants of each of the two groups (default is 'NULL').
a	for 'tweeDE' function, numerical vector with values strictly less than 1. It allows to fix the shape parameter 'a' during differential expression tests for each of the genes (rows of 'object'). for 'testPoissonTweedie' function, numeral value strictly less than 1 which fixes the shape parameter 'a' for the test.
n	maximum number of genes printed.
sort.by	character string, indicating whether genes should be ranked by their P-value (pval), which is the default setting, or by absolute log2 fold-change (log2fc).
log2fc.cutoff	cutoff on the minimum value of the log2 fold change.
pval.adjust.cutoff	cutoff on the maximum adjusted P-value (FDR).
print	logical; it indicates whether the output should be printed on the terminal.
highlight	list of arguments to the points() plotting function in order to highlight genes in the MA or volcano plots. A component called genes is expected to have the identifiers of the genes to be higlighted.
ylab	label on the y-axis of the volcano plot set by default to -log10 of the raw P-value which is what this plot displays on that axis.
x	object returned by the function tweeDE in the case of print and vector of count data in the case of testPoissonTweedie.
saveModel	logical indicating whether the results of fitting the model should be saved or not (default is 'FALSE').
log	logical (default is FALSE). If FALSE, the tested Null Hypothesis states that the difference between the means is 0 while, if TRUE, it states that the quotient between the logarithm of means is equal to 1. For this last case, the standard error is computed using the Delta Method.
...	additional arguments.

Details

'testPoissonTweedie' performs the test for a vector of counts.

'tweeDE' performs the test for a whole 'data.frame'. The P-values are then corrected using the Benjamini and Hochberg method.

Value

'testPoissonTweedie' returns a list with:

'mean': means for each group 'pvalue': p-value for the test

'tweeDE' returns a 'data.frame' with columns

'overallMean': overall mean counts 'meanA': mean counts of the first group 'meanB': mean counts of the second group 'log2fc': logarigthm (base 2) of the fold-change (second group vs. first group) 'pval': p-value for the test 'pval.adjust': adjusted p-value using Benjamini-Hochberg method

References

Esnaola M, Puig P, Gonzalez D, Castelo R and Gonzalez JR (2013). A flexible count data model to fit the wide diversity of expression profiles arising from extensively replicated RNA-seq experiments. BMC Bioinformatics 14: 254

A.H. El-Shaarawi, R. Zhu, H. Joe (2010). Modelling species abundance using the Poisson-Tweedie family. Environmetrics 22, pages 152-164.

P. Hougaard, M.L. Ting Lee, and G.A. Whitmore (1997). Analysis of overdispersed count data by mixtures of poisson variables and poisson processes. Biometrics 53, pages 1225-1238.

See Also

normalizeCounts mlePoissonTweedie

Examples

# Generate a random matrix of counts
counts <- matrix(rPT(n = 1000, a = 0.5, mu = 10, D = 5), ncol = 40)

# Test for differences between the two groups
tweeDE(counts, group = rep(c(1,2),20))

tweeDEseq documentation built on Nov. 8, 2020, 5:59 p.m.