A gene expression matrix of counts with N features and length(n) pairs of samples. The first pair corresponds to the two first column and so on. The total number of samples is 2*n The distribution of the first component (control) is negative binomial with mean mu1 and dispersion parameter 1/size1 (the variance is equal to mu1 + (mu1^2)/size1. For the genes (rows) in gs the distribution of the second component of the pair (case) is a negative binomial distribution with mean mu2 and dispersion 1/size2. For the genes (rows) not in gs (significant genes) the distribution of the second component is negative binomial with mean mu3 and dispersion 1/size3. The matrix outlier contains in the first column the indices of rows and in #' the second column the indices of pairs. For the pairs of counts in outlier the first component follows a negative binomial distribution with mean mu1 and dispersion 1/size1 and the second component follows a negative binomial distribution with mean mu4 and dispersion 1/size4.

rPairedNB(n, N, mu1, size1, mu2, size2, mu3, size3, mu4, size4,
  gs = NULL, outlier = NULL)

Arguments

n

Number of pairs

N

Number of genes

mu1

Mean of the first component

size1

1/size1 is the dispersion of the first component

mu2

Mean of the random count added to the first generated value on the significant genes

size2

1/size2 is the dispersion of the random count added to the first generated value on the significant genes

mu3

Mean of the random count added to the first generated value on the non significant genes

size3

1/size3 is the dispersion of the random count added to the first generated value on the non significant genes

mu4

Mean of the random count added to the first generated value on the outlier genes given in outlier

size4

1/size4 is the dispersion of the random count added to the first generated value on the outlier genes

gs

Significant gene set (a subset of rows in 1:N)

outlier

Matrix (rows as genes and pairs as samples) where the corresponding pair of counts are an outlier gs == NULL correspond with no differential expression