Description Usage Arguments Details Value Author(s) References Examples
Computes the confidence intervals of lambda from bootstrapping
1 |
E1 |
Matrix of k conditions (tissues) and N observations (gene-pair correlations) for rater 1 (experiment 1) |
E2 |
Matrix of k conditions (tissues) and N observations (gene-pair correlations) for rater 2 (experiment 2) |
alpha |
confidence level of CI |
nsam |
number of resamplings with replacements |
mc.cores |
number cores for parallelization |
benchmark |
a logical (default FALSE) that computes the agreement measure for benchmarking experiment 2 (columns) against experiment 1 (rows). |
verbose |
logical. If TRUE it prints the sampling number during computation. |
The function computes the CI of lambda at alpha level from re-sampling with replacement the rows of E1 and E2. E1 and E2 correspond to the observations made by each rater across multiple conditions (columns). Each observation is assumed to be the correlation between two variables.
For gene networks the variables may be the correlation of expression levels between two genes across subjects. In this case, the first column in E1 contains, for instance, all non-redundant pair-wise correlations among the genes in the network for condition (tissue) 1 in experiment 1.
CI |
confidence intervals |
A Caceres
Caceres, A and Gonzalez JR, A measure of agreement across numerous conditions: Assessing when changes in network connectivities across tissues are functional
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 | #simulate E1 and E2 from a multivariate distribution
library(MASS)
library(parallel)
nlevels<-5 #total conditions
ntrue<-3 #number of reliable conditions
cortrue<-0.5 #correlations of reliable conditions
basecor<-0.1 #background correlation
ngenes<-10 #number of genes in network
nonredundantcor<-ngenes*(ngenes-1)/2 #number of non-redundant correlations
#specify covariance matrix for simulation
covmat<-matrix(basecor, ncol=nlevels*2, nrow=nlevels*2)
for(ind in 1:nlevels)
{
covmat[(2*ind-1):(2*ind), (2*ind-1):(2*ind)] <- matrix(c(1,cortrue,cortrue,1), ncol = 2)
if(ind > ntrue)
covmat[(2*ind-1):(2*ind), (2*ind-1):(2*ind)] <- matrix(c(1,basecor,basecor,1), ncol = 2)
}
simul <- mvrnorm(nonredundantcor, mu = rep(0,nlevels*2), Sigma = covmat, empirical = TRUE)
###
#45 non-redundant gene correlation pairs from a 10 gene-network on 5 conditions, 3 of which are inter experiment reliable (3/5=0.6)
E1 <- simul[,seq(1,ncol(simul),2)] #experiment 1
E2 <- simul[,seq(1,ncol(simul),2)+1] #experiment 2
CIlambda(E1=E1, E2=E2, alpha=0.05, nsam=100, mc.cores=2)
|
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.