iFad (version 3.0)

iFad-package: An integrative factor analysis model for drug-pathway association inference

Description

This package implements a bayesian sparse factor model for the joint analysis of paired datasets, one is the gene expression dataset and the other is the drug sensitivity profiles, measured across the same panel of samples, e.g., cell lines. Prior knowledge about gene-pathway associations can be easily incorporated in the model to aid the inference of drug-pathway associations.

Details

Package:

iFad

Type:

Package

Version:

3.0

Date:

2014-03-25

License:

GPL (version 2 or later)

LazyLoad:

yes

install.packages("iFad")

Examples

library(Rlab)
library(MASS)
library(coda)
library(ROCR)

#Simulate datasets

data_simulation(K=10,G1=30,G2=30,J=15,eta0=c(0.2,0.2),
eta1=c(0.2,0.2),density=c(0.1,0.1),alpha_tau=1,
beta_tau=0.01,SNR=0,file_name="demo_data.RData")


#Gibbs sampling

data(matrixY1)
data(matrixY2)
data(matrixL1)
data(matrixL2)


gibbs_sampling(matrixY1, matrixY2, matrixL1, matrixL2, 
eta0=c(0.2,0.2), eta1=c(0.2,0.2), alpha_tau = 1, 
beta_tau = 0.01, tau_sig = 1, max_iter = 5, 
thin = 1, file_name="Demo_Gibbs_result.RData")


#Traceplot
data(tau_g_chain)
mcmc_trace_plot(tau_g_chain,plot_file_name="Demo_traceplot.pdf", 
index=1:10)

#ROC plot
data(matrixZ1)
data(matrixZ2)
data(matrixZ_chain)
ROC_plot(matrixZ1, matrixZ2, matrixZ_chain, plot_name="ROC_plot.pdf", 
result_file_name="ROC_result.RData", burn=1)

#RMSE plot
data(Y1_mean)
data(Y2_mean)
data(matrixY1)
data(matrixY2)
data(matrixZ_chain)
data(matrixW1)
data(matrixW2)
data(matrixW_chain)
data(matrixX)
data(matrixX_chain)
Ymean_compare(Y1_mean,Y2_mean,matrixY1, matrixY2, matrixZ_chain, 
matrixW1, matrixW2, matrixW_chain, matrixX, matrixX_chain, 
result_file_name="RMSE_demo.RData", plot_name="RMSE_plot.pdf")