Skip to content

Latest commit

 

History

History
131 lines (84 loc) · 4.25 KB

README.md

File metadata and controls

131 lines (84 loc) · 4.25 KB
Raw

Contains Clarity scripts that use the clarity-ext framework at https://github.com/molmed/clarity-ext

clarity-ext

Usage:

  • install-req.sh (only when setting up)
  • Activate your virtual environment (must be python 2)
  • setup.sh

These scripts require the libraries clarity-ext and its version of genologics. Both are included as submodules, and will be updated by calling setup.sh.

Special note for covid19: In order to update these requirements for now, make PRs into the branch covid on:

https://github.com/Molmed/clarity-ext
https://github.com/Molmed/genologics

and make sure to git add these commit references in this repository.

Releasing to production:

Releasing to the productions servers concists of two major steps. Pushing the script changes, and pushing the config changes. The first one is carried out by:

  1. Merging the dev branch to the prod branch. When this has happened, these changes should within a minute be pushed to the production server.
  2. Creating a release on the prod branch with accompanying release notes, using the Github web interface.

For details on how to release the configuration changes see "Exporting configuration".

Exporting configuration:

There's a helper script for exporting in the ./deployment directory:

Get dependencies

First call

./get-slice-tool.sh <server>

in order to get the required jar file into your ./bin directory

Add configuration

Create a yaml configuration file at ~/.slices.config

This configuration file should contain information on this format:

staging:
  server: server
  username: username
  password: password
prod:
  server: server
  username: username
  password: password

Create a manifest file

Then you can run the slices tool like this:

python ./slices.py manifest staging  # Creates a new manifest file in your export directory

Copy this manifest file over the file deployment/manifest.txt and remove everything that shouldn't be imported to production.

Create a package

Run this to create a package:

python ./slices.py export staging

This will create a new xml package in the exports directory, which can be imported into production.

Import a package into production

First call this:

python ./slices.py import prod --validate <package file>

Read through the results and ensure that there are no warnings or errors. If ready, call:

python ./slices.py import prod <package file>

Save the history

zip import-v1-prod.zip configslicer.log exports/*
scp import-v1-prod.zip glsai@ctmr-lims-stage:/opt/gls/clarity/users/glsai/deployment/history/

Import samples format

Samples are imported by uploading a CSV file to the step Covid19 Create samples. First, add a control to the step. It will not be used, it's just there because Clarity requires a sample in order to start a workflow.

We assume that the samples are in a plate.

The CSV file must have these fields:

* name: Name of the sample
* well: The alphanum index of the sample (e.g. A1 or A:1)

Example:

name,well
sample1,A1
neg_rna,A2

Furthermore, the research engineer must provide the following step UDFs:

* Assign to workflow: The name of the workflow to which the samples should be added
* Project: The name of the project

Other step UDFs are not required as they have defaults, but can be set e.g. for test purposes:

* Sample name: The format of the sample name. The user may specify the dynamic fields:
    * {name}: The original name in the CSV
    * {date:<python format>}: The timestamp when the test started, formatted as a string according to
      python date and time format specifiers which are documented [here](https://strftime.org/)
    * {index}: The index of the sample in the batch (running number)
* Control name: The format of the control name. Provides the same fields as Sample name.
* Container name: Format of the container name. Can use these fields:
    * {name}: Sampe as for Sample name
    * {date:<python format>}: Same as for Sample name
* Control mapping: A comma separated list of mappings from a pattern that matches controls in the name field in the CSV
  and returns the value that should be added to the UDF "Control type".
* Container type: Defaults to 96 well plate