Import data using the Command Line Interface (CLI)

Description

This chapter will show how to import data for another user, using Command Line Interface (CLI).

The import can be done by any user as long as they import the data for themselves.

In case of the import for others, the user importing the data needs to have some admin (or restricted-admin) privileges. More information about restricted privileges can be found at restricted-admins.

In the example workflow below, a user with restricted administrator privileges is used with login name importer1. This could be in real life e.g. a facility manager.

We will show:

  • How to import data using the CLI for myself and for others into a remote OMERO.server
  • How to import data using the CLI “in-place”, which means not copying the imported data into OMERO. Instead, OMERO will point to the original location of “in-place” imported files, thus preventing data duplication.
  • How to deal with imports of large amounts of data in CLI, using the –bulk option and helper csv and yml files which define what is to be imported and how.

Setup

CLI Importer installation

Client libraries from the OMERO.server have to be installed on the client to import images using CLI.Please read the installation instructions.

Note: When importing for another user using the CLI, the importer1 does not have to be a member of the target group.

Step-by-step

  1. If you did not do so already, open a terminal window on your local machine and activate the conda environment where your omero-py is installed (see the Setup section above):

    $ conda activate myenv

  2. Set the OMERODIR variable to point to the downloaded and unzipped OMERO server dir (see the Setup section above):

    $ export OMERODIR=/path/to/OMERO.server-x.x.x-ice36-bxx

  3. Login to the OMERO.server you wish to import to. This can be a remote OMERO.server. In the below example we log in as the user importer1, who will import for themselves:

    $ omero -s your-server-address -u importer1 login

  4. Go to the directory where there are some images you wish to import:

    $ cd /path/to/images/directory/

  5. Create an OMERO Dataset to import to:

    $ DID=$(omero obj new Dataset name=import_for_myself)

  6. Import the images as importer1 into the newly created Dataset. You can import a single image as in the example below, or a whole directory of images:

    $ omero import -d $DID <your-image-name>

  7. Log in to OMERO.web as importer1 and verify the imported image in the newly created dataset import_for_myself.

  8. Import an image for another user, for example user-1. For that, your importer1 user logs in as user-1:

    $ omero --sudo importer1 -u user-1 login

  9. Create a Dataset import_for_user_one as user-1:

    $ DID=$(omero obj new Dataset name=import_for_user_one)

  10. Import the data in the newly created Dataset:

    $ omero import -d $DID <path-to-image-or-directory-with-images>

  11. Check that the image(s) are successfully imported and that the image(s) and the containing Dataset both belong to user-1 (not importer1).

In-place Import using the CLI

Instead of being copied into OMERO’s managed repository, the image files stay at their original place and are just linked into the repository.

It is only available for the CLI importer, using the argument --transfer=ln_s.

../../_images/importcli1.png

Advantages:

  • All in-place import scenarios provide non-copying benefit. Data that is too large to exist in multiple places, or which is accessed too frequently in its original form to be renamed, remains where it was originally acquired.

Limitations:

  • Only available on the OMERO server system itself.
  • Do not edit or move the files after an in-place import. OMERO may no longer be able to access them if you do.

Important:

A user performing an in-place import MUST have:

  • a regular OMERO account

  • an OS-account with ability to run omero commands on server machine

  • read access to the location of the data

  • write access to the ManagedRepository or one of its subdirectories. Please check the ManagedRepository documentation.

    ../../_images/importcli2.png

Step-by-step:

  1. Connect to the machine on which the OMERO.server is running as OS user importer1 using ssh.

  2. The aim is to import an image from /OMERO/in-place-import/FRAP:

    $ ls /OMERO/in-place-import/FRAP

  3. Activate the virtual environment where omero-py is installed or add it to PATH. In the example below, the path to the OMERO.server is /opt/omero/server:

    $ export PATH=/opt/omero/server/venv3/bin:$PATH

  4. Point OMERODIR to the location where the OMERO server is installed e.g.:

    $ export OMERODIR=/opt/omero/server/OMERO.server

  5. Import now data for another user, this time a large image where the advantage of not copying the image file onto the server is most visible. The importer1 user logs in as user-1:

    $ omero --sudo importer1 -u user-1 login

  6. Create a Dataset import_for_user_one:

    $ DID=$(omero obj new Dataset name=import_for_user_one)

  • ‘In place’ import a large SVS file into the import_for_user_one dataset:

    $ omero import -d $DID --transfer=ln_s /OMERO/in-place-import/svs/77917.svs

  • Check that the image is successfully imported.

  • Click on the paths icon image3 to show the difference between the normal and in-place (ln_s) imported images. Validate that In-place import is indicated image4.

  • Note: The script in_place_import_as.sh shows how to perform the in-place import steps described above in one single command.

Bulk Import using the CLI

In this example, we show how to combine several import strategies using a configuration file. This is a strategy heavily used to import data to IDR.

We import two folders named siRNA-HeLa and condensation.

Note: Connecting over SSH is necessary only if you intend to import in-place. If you do not wish to perform the bulk import in “in-place” manner, you can connect to the server remotely using locally installed OMERO.cli and adjust the bulk.yml file by commenting out the transfer... line, then follow the steps as described below.

  1. Open a terminal and connect to the server (for example as importer1) over SSH. Alternatively, use your local terminal with installed OMERO.cli if not importing “in-place”.

  2. Description of the files used to set up the import (see bulk.yml, import-paths.csv and import-bulk.html#bulk-imports for further details).

    • import-paths.csv: (.csv, comma-separated values) this file has at least two columns. In this case the columns are separated by commas. The first column is the name of the target Dataset and the second one is the path to the folder to import. We will import two folders (the import-paths.csv has two rows).

      Example csv (note the comma between the “HeLa” and “/OMERO…”):

      *Dataset:name:Experiment1-HeLa,/OMERO/in-place-import/siRNAi-HeLa*

      *Dataset:name:Experiment2-condensation,/OMERO/in-place-import/condensation*

    • bulk.yml: this file defines the various import options: transfer option, checksum algorithm, format of the .csv file, etc. Note that setting the dry_run option to true allows to first run an import in dry_run mode and copy the output to an external file. This is useful when running an import in parallel. Comment out the transfer" "ln_s" if not importing “in-place”.

      Example bulk.yml:

    *continue: "true"*

*transfer: "ln_s"*

*# exclude: "clientpath"*

*checksum_algorithm: "File-Size-64"*

*logprefix: "logs"*

*output: "yaml"*

*path: "import-paths.csv"*

*columns:*

    -  *target*

    -  *path*

  3. Activate the virtual environment where omero-py is installed or add it to PATH. In the example below, the path to the OMERO.server is /opt/omero/server:

    $ export PATH=/opt/omero/server/venv3/bin:$PATH

  4. Point OMERODIR to the location where the OMERO server is installed e.g.:

    $ export OMERODIR=/opt/omero/server/OMERO.server

  5. Find the place where the bulk.yml file is located, for example /OMERO/in-place-import:

    $ cd /OMERO/in-place-import

  6. The importer1 (Facility Manager with ability to import for others) OMERO user logs in as user-1:

    $ omero --sudo importer1 -u user-1 login

  7. Import the data using the --bulk command:

    $ omero import --bulk bulk.yml

  8. Go to the webclient during the import process to show the newly created dataset. The new datasets in OMERO are named Experiment1-HeLa and Experiment2-condensation. This was specified in the first column of the import-paths.csv file.

  9. Select an image.

  10. In the right-hand panel, select the General tab to validate:

    • Click on image3 to show the import details.
    • Validate that In-place import is indicated image4 in case you imported “in-place”.

Advantages:

  • Large amount of data imported using one import command.
  • Heterogeneous data for multiple users can be imported using bulk import in combination with bash scripting, e.g. in_place_import_as.sh
  • Reproducible import.

Limitations:

  • Preparation of the .csv or .tsv file.

Combine the CLI imports with post-import steps

The following example shows how to do the import on CLI and follow-up operations like rendering and metadata import in one step.

In many cases, the rendering and metadata import is best done separately, as the visual checking of the imported images might be crucial for further rendering and metadata import, see Change image rendering settings and channel names using the Command Line Interface (CLI) and Import metadata using the Command Line Interface (CLI) for details on this.

Further, the images you are importing might need a range of different rendering settings, not just one set of settings for all of them. Also for this case, the step-by-step approach, first importing the images, only then deciding on the rendering strategy and preparing the renderingdef.yml files, is preferable.

Nevertheless, there are cases which do not need visual checks and use a single rendering for all images, for which a streamlined sequence of commands is offered below which will perform all three steps (import, rendering and metadata import) in one single session on the CLI.

Step-by-step

  1. If you did not do so already, open a terminal window on your local machine and activate the conda environment where your omero-py is installed (see the Setup):

    $ conda activate myenv

  2. Set the OMERODIR variable to point to the downloaded and unzipped OMERO server dir (see the Setup):

    $ export OMERODIR=/path/to/OMERO.server-x.x.x-ice36-bxx

  3. Prepare an renderingdef.yml file, by either creating a new one or downloading https://raw.githubusercontent.com/ome/training-scripts/master/maintenance/preparation/renderingdef.yml.

  4. Prepare an annotation.csv file, by creating a new file or downloading the provided example file. In the example below, we use the file simple-annotation.csv. The Dataset names in this CSV file must match the Dataset names in OMERO as created in the DID variable definition line in the command below. The Image names in the CSV file must match the file names in your imported folder.

  5. Prepare a bulkmap-config.yml file. In the example below, we use the file simple-annotation-bulkmap-config.yml.

  6. Log in to the OMERO.server you wish to import to. This can be a remote server if you do not wish to import in-place.

  7. Import, render and annotate in a single command sequence below:

    $  PID=$(omero obj new Project name='Project_import_concatenate')
$  DID=$(omero obj new Dataset name='siRNAi-HeLa')
$  omero obj new ProjectDatasetLink parent=$PID child=$DID
$  omero import -d $DID /path/to/data/folder/or/image/siRNAi-HeLa --file import.out
$  omero render set $DID renderingdef.yml
$  omero metadata populate --report --batch 1000 --file /path/to/downloaded/simple-annotation.csv $PID
$  omero metadata populate --context bulkmap --cfg simple-annotation-bulkmap-config.yml --batch 100 $PID

  8. Log in to OMERO.web and check that the images are imported, have the expected rendering setttings and also the annotations in the form of Key-Value pairs on each imported image.

For more information about CLI import options, go to import.html.