Is it possible to use multi-threaded work to increase the speed of calculations?

[baoyulinn]

Dear professor
Is it possible to use multi-threaded work to increase the speed of calculations?

[Ali-mhm]

Dear professor Is it possible to use multi-threaded work to increase the speed of calculations?

hi, you can use these line of code to increase number of CPUs' cores

library(MRPRESSO)
library(parallel)
library(data.table)

# Load your summary statistics from a file
t<-fread("C:/Users/pc/Downloads/Compressed/presso.csv")
t<-t[!duplicated(t$SNP),]
snp<-as.data.frame(t$SNP)
t<-t[,-7]
t<-as.data.frame(t(t))
colnames(t)<-snp$`t$SNP`
t<-as.data.frame(t(t))
# Replace 'your_file_path.txt' with the actual path to your file
SummaryStats <- t

# Set the number of cores you want to use
num_cores <- 20

# Create a cluster with the specified number of cores
cl <- makeCluster(num_cores)

# Export necessary objects and libraries to the cluster
clusterExport(cl, c("SummaryStats", "mr_presso"))
clusterEvalQ(cl, library(MRPRESSO))

# Define a function to run mr_presso
run_mr_presso <- function(i) {
  mr_presso(BetaOutcome = "Y_effect", 
            BetaExposure = "E1_effect", 
            SdOutcome = "Y_se", 
            SdExposure = "E1_se", 
            OUTLIERtest = TRUE, 
            DISTORTIONtest = TRUE, 
            data = SummaryStats, 
            NbDistribution = 6000,  
            SignifThreshold = 0.05)
}

# Run the function in parallel
results_list <- parLapply(cl, 1:num_cores, run_mr_presso)

# Stop the cluster
stopCluster(cl)

i hope this let you to increase speed of calculation