Function to simulation study

justDo2(n, N, nsig, noutlier = c(1, 1), theta, model = c("Poisson",
  "NB"), test = proportion.p, norness = 50, nsim = 100,
  method = "dbinom", is.decreasing = FALSE, proc = c("owb", "owc",
  "er", "de"), do.auc = FALSE)

Arguments

n

Number of pairs

N

Number of genes

nsig

Number of significant genes (rows (noutlier+1):(noutlier+nsig)

noutlier

Vector giving the number of outlier genes noutlier[1] (rows 1:noutlier[1] in the expression matrix) and number of pairs (columns 1:noutlier[2])

theta

Vector with the lambda values for tami::rPairedPoisson or matrix with two columns and 4 rows where the entry theta[i,1] is the mean and 1/theta[i,2] the dispersion for the i-th random variable

model

Distribution of the random counts: "Poisson" or "NB" for Poisson or negative binomial distribution

test

Function comparing counts within each pair

norness

Number of values for orness if orness == NULL

nsim

Number of simulations

method

Method to generate the orness values c("dbinom","fuller"). See owa.weights.

is.decreasing

The values to be aggregated in decreasing order

proc

Procedures to be tested: owb OWA with between-pair distribution, owc OWA with complete distribution, er edgeR, de DESEq2.

do.auc

Is the AUC value to be calculated?