I have been trying to conduct a classic GSEA using the broad institute software and can not figure out how to compile it in R for the life of me. I am aware that there exists other packages in R that are almost the same, such as fgsea
however, I would like to use this one first and then compare it with these other methods.
I have already generated a ranked list for my contrast of interest and save it as a .rnk
file. Here is what I tried to do:
gseaCallH <-
paste("java -Xmx2G -cp C:/Users/omk/Desktop/BIOINFORMATIC/MsigDB/gsea2-2.2.4.jar xtools.gsea.GseaPreranked -param_file C:/Users/omk/Desktop/BIOINFORMATIC/MsigDB/GSEAPrerankedParameters.txt -rnk C:/Users/omk/Desktop/BIOINFORMATIC/MsigDB/gsea_treatFU_min.rnk -gmx C:/Users/omk/Desktop/BIOINFORMATIC/MsigDB/h.all.v6.2.symbols.gmt -rpt_label gsea_treatFU_min.rnk -out C:/Users/omk/Desktop/BIOINFORMATIC/MsigDB/C7",sep="")
testGSEA <- system(command = gseaCallH, intern = TRUE, ignore.stderr = TRUE)
When I run the system
command I get the following error:
had status 1
At first I thought maybe it had something to do with JAVA not being properly installed however, I have installed it and still nothing. Maybe the installation was not done correctly? I have tried running it directly from cmd (I'm running windows 10) and got the following error:
Error occurred during initialization of VM
Could not reserve enough space for 2097152KB object heap
Can anyone advise on how to proceed?
R CMD javareconf
) maybe it helps too. Because the way you run it now from R is sending it back to the command line, so you should get the error from the command line $\endgroup$