Child pages
  • How to Upload Studies To ScanBase
Skip to end of metadata
Go to start of metadata

When a pydpiper pipeline has been completed and the results are ready to go into our ScanBase database or OBI's brainCODE database you will need to upload your results into the ScanBase googlesheet.To simplify this process there is an R package "scanbaseTools" installed on the system that allows you to upload easily.

At the time of writing, scanbaseTools needs to compute extra determinants before the study is ready to be uploaded. It is recommended that you run it over night on your machine. This will be resolved in future releases of pydpiper.

Step-by-step guide

  1. You are required to prepare some metadata before you can begin uploading. Presently the metadata consists of 4 small csvs. This is where you give all the details needed to understand the study. Please try to be as complete as possible with the metadata.
    1. Individual scan metadata: example_scan_sheet_setd5.csv

    2. Genotypes used in the study:example_genotypes_sheet_setd5.csv

    3. Treatments used in the study:example_treatments_sheet_setd5.csv

      These files should be up to date, but to be sure you can also use the versions found in scanbaseTools by running

      Rscript -e 'list.files(system.file(package = "scanbaseTools"), pattern = ".csv$")'
    The scan csv should have number of lines equal to the number of input scans, the study csv should have a single line, the genotypes should have lines equal to the number of genotypes in your study, same for treatments. Talk to Jacob if you are in doubt about any of the sheets or columns.

  2. Be prepared for a long running command. Possibly sign in and do this on HPF (whoops, not yet, the required version of minc-stuffs isn't installed there yet).
  3. scanbaseTools will prompt you to log in to googlesheets, you'll need to use the account mouse.imaging.centre@gmail.com, password available from me and Jacob on request.
  4. module load minc-stuffs/0.1.24
    R
    library(scanbaseTools)
    upload_study(ppd = <path to where you ran pydpiper>,
                 scan = <path to scan sheet>,
                 study = <path to study sheet>,
                 treatments = <path to treatments sheet>,
                 genotypes = <you get it>)
  5. Be patient.
  6. ScanbaseTools will attempt to upload the data to google sheets.
  7. Check that the upload occurred successfully by inspecting test_scanbase_40um on googlesheets.
  8. If you were unable to upload and the command fails, don't fret, just run the `upload_study` command again. It should be much much faster the second time.
  9. Once we've all determined that the uploads are working properly, I'll switch the default to upload to scanbase_40um (not the test version).