huthvincent / cGLMM

source code of Privacy-preserving Construction of Generalized Linear Mixed Model for Biomedical Computation

cGLMM

Implementation of the distributed version of cGLMM with Homomorphic encryption in R

Non-distributed Version of GLMM

muti_fe_latest.Rmd

Distributed Version

Generate simulated data

Run Generate_test_data.r to generate two files stands for two clients

Install requried packages

require(tcltk) require(svMisc) require(svSocket) library(gmp) library(homomorpheR)

Run instruction

1. Lanch Rstudio and set the same fold as working direcotry for both client (Fed_GLMM_client_function.r) and server (Fed_GLMM_server.r).
2. Lanch at least 3 session stand for server, client1, and client2 in Rstudio
3. At each session, run the corresponding scripts at runFed_GLMM.r

Run instruction(For the HE version)

1. Run initial_parameters.R to save initial parameters as initial_parameters.RData
2. Run generate_keyPair.R to save Homomorphic Key Pair as priv_key.RData, pub_key.RData
3. Lanch Rstudio and set the same fold as working direcotry for both client (Fed_GLMM_client_function.r) and server (Fed_GLMM_server.r).
4. Lanch at least 3 session stand for server, client1, and client2 in Rstudio
5. At each session, run the corresponding scripts at runFed_GLMM.r

Ref:

Introduction to Homomorphic Computation in R# cGLMM