generateExpressionData: Generate Gene Expression Data

In survGenesInterim: Simulation of Survival Studies with Microarray Data

Description

This function simulates gene expression data based on the multivariate normal distribution for two groups of samples.

Usage

  generateExpressionData(fc = rep(0, 100),
    Sigma.1 = diag(100), Sigma.2 = NULL, N.1 = 10,
    N.2 = 10, use_cholesky = FALSE)

Arguments

fc	the vector of foldchanges between the two groups
Sigma.1	the covariance matrix describing the correlation between the genes in group one
Sigma.2	the covariance matrix describing the correlation between the genes in group two. If this is NULL, the case of equal covariances is assumed and Sigma.2 is set to Sigma.1.
N.1	the sample size of group one
N.2	the sample size of group two
use_cholesky	this is a boolean parameter that indicates whether the covariance matrices are cholesky decomposed. This is an enourmous speed up when simulating.

Value

X.1	the simulated gene expression levels of group one
X.2	the simulated gene expression levels of group two
d	the dimension, i.e. the number of genes
fc	the fold change vector. This is the unchanged parameter to the function.

Author(s)

Andreas Leha [e-mail]

See Also

generatePatientData

Examples

## create a vector of fold changes
fc <- c(rep(0, 500), ceiling(rnorm(500, 0, 1)-0.5))

## uncorrelated genes
Sigma.1 <- diag(1000)
Sigma.2 <- diag(1000)

## the sample sizes
N.1 <- 30
N.2 <- 30

## Not run: G <- generateExpressionData(fc, Sigma.1, Sigma.2, N.1, N.2)

survGenesInterim documentation built on May 31, 2017, 3:21 a.m.

Simple Estimation of the False Discovery Rate

survGenesInterim index

Generate Patient Data