survGenesInterim-package: Simulation of Survival Studies with Microarray Data
In survGenesInterim: Simulation of Survival Studies with Microarray Data

Description Details Author(s) References Examples

The main part of this package is the simulation of a survival study based on simulated gene expression level data and simulated patient data. Functions to generate such gene expression level data and patient data are included as well. Resulting error rate (FDR) and power rate can be visualized. Thus, this package can be used for sample size or power estimations in the planning of a study.

Package:	survGenesInterim
Type:	Package
Version:	1.0
Date:	2010-07-19
License:	GPL (>= 2)
LazyLoad:	yes

A frequent objective of clinical studies is to detect genes or biomarkers that can predict the outcome of therapy and thus the survival of patients. Therefore, gene expression levels in samples of pretherapeutic ill or healthy samples are measured by DNA microarrays and compared to survival data of the patients. Because usually, tissue samples are only available at distinguished time points and it can take several years until the full planned sample size is available and follow-up is complete. In such long-lasting studies it is beneficial to obtain interim results already before their planned end. The details are of the simulation are given in Leha et al. (2010).

This package allows to simulate such situation and to visualize the results, which is useful during the planning of this kind of studies, i.e. to assist in sample size planning.

Andreas Leha andreas.leha@med.uni-goettingen.de

Leha, Andreas and Beissbarth, Tim and Jung, Klaus (2010): "Sequential Interim Analyses of Survival Data in Microarray Experiments", submitted

fc <- c(rep(0, 500), ceiling(rnorm(500, 0, 1)-0.5))
Sigma.1 <- diag(1000)
Sigma.2 <- diag(1000)

N <- 50
l.1.tick <- 60
l.2 <- 60
lambda <- 60

M.1 <- 2
M.2 <- 2
alpha <- 0.05
powerThreshold <- 0.8
adjustment <- "BH"

numSimRuns <- 2

## Not run: result <- interimTrialSimulation(fc, Sigma.1, Sigma.2,
                                          N, l.1.tick, l.2, lambda,
                                          M.1, M.2, alpha, powerThreshold, adjustment,
                                          numSimRuns)
## End(Not run)