Description
Quantify and Control Reproducibility in High-Throughput Experiments.
Description
Estimate the proportions of the null and the reproducibility and non-reproducibility of the signal group for the input data set. The Bayes factor calculation and EM (Expectation Maximization) algorithm procedures are also included.
README.md
Installation
library(devtools)
install_github("ArtemisZhao/INTRIGUE/R_src")
Example dataset
library(INTRIGUE)
data("heterodata")
This is a simulated datasets with n=5000 experimental units (e.g., genes) and m=2 replicates. The true proportions for the null, the reproducible and the irreproducible signal categories are 0.80, 0.18, 0.02, respectively.
CEFN prior
##for convenience, only test on first 100 units
hetero.out.cefn<-hetero(heterodata[1:100,],fdr.level=0.05)
META prior
hetero.out.meta<-hetero(heterodata[1:100,],use_cefn=FALSE,fdr.level=0.05)
Outcome
names(hetero.out.cefn)
## overall proportion
print(hetero.out$est_prop)
Return a list containing the following,
gridweight: estimated weight on each grid point.
ind_prob: individual probability of falling into three groups.
est_prop: estimated overall proportion for three groups.
lfdr: false discovery rate.
significant: decision based on the fdr.level in the input.