tweeDEseq (version 1.18.0)

compareCountDist: Compare count data distributions

Description

Compares the empirical and estimated distributions for different count data models

Usage

compareCountDist(x, plot=TRUE, ...)

Arguments

x

numeric vector containing the read counts.

plot

If TRUE (the default) then the plot with the ECDF function for the counts and the three different Poisson-Tweedie distributions is produced, otherwise no graphical output is given and this only makes sense if one is interested in the returned value (see value section below).

...

Further arguments to be passed to the plot function.

Value

List with the following components:

a

shape parameter estimated from the input data x.

p.value

P-value for the test that the data follows a negative binomial distribution, i.e., $H_0:a=0$.

Details

This function serves the purpose of comparing a empirical distribution of counts with three Poisson-Tweedie distributions arising from estimating mean, dispersion and setting $a=1$ for comparing against a Poisson, $a=0$ for comparing against a negative binomial and estimating the shape parameter a from data too. The legend shows the values of the $a$ parameter and the P-value of the likelihood ratio test on whether the expression profile follows a negative binomial distribution ($H_0:a=0$).

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

See Also

qqchisq testShapePT

Examples

# Generate 500 random counts following a Poisson Inverse Gaussian
# distribution with mean = 20 and dispersion = 5
randomCounts <- rPT(n = 500, mu = 20, D = 5, a = 0.5)

xx <- compareCountDist(randomCounts, plot=FALSE)
xx