Skip to content

Latest commit

 

History

History
35 lines (32 loc) · 1.59 KB

README.md

File metadata and controls

35 lines (32 loc) · 1.59 KB
Raw

MTVR

R version of MultiTrans (http://genetics.cs.ucla.edu/multiTrans/)

Prerequistes for running Source Code

(suggest using Linux environment)

Usage

Step 1. Install and load package

library(devtools)
install_github("2eding/MTVR")
library(MTVR)

Step 2. Input data

X <- as.matrix(data.table::fread("X_rightdim.txt")) (individual x SNP)
Y <- as.matrix(data.table::fread("Y.txt")) (Phenotype x individual)

  • data.table::fread <= This package is useful for input large data
  • X_rightdim.txt is snp data
  • Y.txt is gene expression data

Step 3. Calculate the kinship

K <- Kinship(X, outPath, outName)

Step 4. Calculate the variance components

VC <- varComp(Y, K, numThreads, outPath, outName) (numThreads = Determines how many threads to use)

Step 5. Estimate correlation in the rotated space

corrMatrix <- generateR(X, K, VC, outPath, outName)
covBandMatrix <- generateC(windowSize, corrMatrix, outPath = "./", outName = "c.txt")

Step 6. Run SLIDE

step1 <- system.file('./slide_1prep', package='MTVR')
step2 <- system.file('./slide_2run', package='MTVR')
step3 <- system.file('./slide_3sort', package='MTVR')
step4 <- system.file('./slide_4correct', package='MTVR')

system(paste(step1, "-C", "covBandMat path", windowSize, "outPath & outName"))
system(paste(step2, "step1's outPath & outName", "outPath & outName", simulationNum, seedNum))
system(paste(step3, "outPath & outName", "step2's outPath & outName"))
system(paste(step4, "-p", "step2's outPath & outName", "threshold file path", "outPath & outName"))