gaga (version 2.18.0)

simGG: Prior predictive simulation

Description

simGG simulates parameters and data from the prior-predictive of GaGa/ MiGaGa models with several groups, fixing the hyper-parameters.

simLNN simulates from a log-normal normal with gene-specific variances (LNNMV in package EBarrays). simNN returns the log observations.

Usage

simGG(n, m, p.de=.1, a0, nu, balpha, nualpha, equalcv = TRUE, probclus
= 1, a = NA, l = NA, useal = FALSE)
simLNN(n, m, p.de=0.1, mu0, tau0, v0, sigma0)
simNN(n, m, p.de=0.1, mu0, tau0, v0, sigma0)

Arguments

n

Number of genes.

m

Vector indicating number of observations to be simulated for each group.

p.de

Probability that a gene is differentially expressed.

a0, nu

Mean expression for each gene is generated from 1/rgamma(a0,a0/nu) if probclus is of length 1, and from a mixture if length(probclus)>1.

balpha, nualpha

Shape parameter for each gene is generated from rgamma(balpha,balpha/nualpha).

equalcv

If equalcv==TRUE the shape parameter is simulated to be constant across groups.

probclus

Vector with the probability of each component in the mixture. Set to 1 for the GaGa model.

a, l

Optionally, if useal==TRUE the parameter values are not generated, only the data is generated. a is a matrix with the shape parameters of each gene and group and l is a matrix with the mean expressions.

useal

For useal==TRUE the parameter values specified in a and l are used, instead of being generated.

mu0,tau0

Gene-specific means arise from N(mu0,tau0^2)

v0, sigma0

Gene-specific variances arise from IG(.5*nu0,.5*nu0*sigma0^2)

Value

Object of class 'ExpressionSet'. Expression values can be accessed via exprs(object) and the parameter values used to generate the expression values can be accessed via fData(object).

Details

For the GaGa model, the shape parameters are actually drawn from a gamma approximation to their posterior distribution. The function rcgamma implements this approximation.

References

Rossell D. (2009) GaGa: a Parsimonious and Flexible Model for Differential Expression Analysis. Annals of Applied Statistics, 3, 1035-1051.

Yuan, M. and Kendziorski, C. (2006). A unified approach for simultaneous gene clustering and differential expression identification. Biometrics 62(4): 1089-1098.

See Also

simnewsamples to simulate from the posterior predictive, checkfit for graphical posterior predictive checks.

Examples

#Not run. Example from the help manual
#library(gaga)
#set.seed(10)
#n <- 100; m <- c(6,6)
#a0 <- 25.5; nu <- 0.109
#balpha <- 1.183; nualpha <- 1683
#probpat <- c(.95,.05)
#xsim <- simGG(n,m,p.de=probpat[2],a0,nu,balpha,nualpha)
#
#plot(density(xsim$x),main='')
#plot(xsim$l,xsim$a,ylab='Shape',xlab='Mean')