How to use R-based PLINK plug-in in RAP?

I am trying to use R on PLINK via R-based PLINK plug-in. In order to make the R plug-in function work, Rserve must be running in the background before invoking the PLINK command. I typed the full path of where Rserve was installed in the command line of Swiss Army Knife: /usr/local/lib/R/bin/R CMD /home/rstudio-server/R/x86_64-pc-linux-gnu-library/4.1/Rserve/libs//Rserve However, it was not working. I'm wondering if anyone know how to make this work?   I would appreciate any and all feedback.

Comments

5 comments

  • Comment author
    Chai Fungtammasan DNAnexus Team

    We don't have the view access for your analysis, so you may need to contact ukbiobank-support@dnanexus.com for debugging question. I check in swiss-army-knife, but didn't find path you mentioned, so I guess you installed them yourselves.

    There is a plink package in R. We have Rstudio on UKB-RAP which you can run multiple programming language there. I would personally go with that route.

     

    0
  • Comment author
    Former User of DNAx Community_26

    Thanks! Do you know the name of the plink package in R?

    0
  • Comment author
    Chai Fungtammasan DNAnexus Team

    My bad, the PLINK package I found is for something else.

    However, you should still be able to install program you need to do interactive work in Rstudio though.

     

    0
  • Comment author
    Former User of DNAx Community_26

    I have contacted  ukbiobank-support@dnanexus.com but their feedback was not helpful. I've tried everything I could but still cannot do interactive work on PLINK with R. When using any R-based PLINK plug-in, Rserve must be running in the background before invoking the PLINK command. And the main problem is that I cannot set up the Rserve package when I am using swiss-army-knife.

     

    0
  • Comment author
    Chai Fungtammasan DNAnexus Team

    sorry to hear. Okay. let's try this from scratch.

    1) run one of interactive workstation. This could be ttyd, cloud_workstation, Rstudio, Jupyternotebook, etc.

    2) using terminal of that app and to install PLINK https://www.cog-genomics.org/plink/

    3) same as #2 but install Rserver

    4) run Rserver in background and see if you can run PLINK plug in

    5) If this is working you can try to encapsulate this environment for future usage. For example, in cloud_workstation, you can use dx-create-snapshot to save the workspace and load it for future usage. Or if you are familiar with how to create docker image, you can wrap it up as Docker and run it in non-interactively.

     

    Would this approach works for you? It may look cumbersome, but all those installation is just one time and you can reuse your snapshot for the future.

     

    0

Please sign in to leave a comment.