<aside> 🔬
wrmXpress is a software suite that analyzes plate-based imaging data generated on our microscopy systems. It combines bespoke modules with pipelines developed within other frameworks, and can be deployed locally or distributed on a high-throughput computing network. See the wrmXpress GitHub repository for details.
</aside>
Below are lab instructions on how to deploy wrmXpress on the CHTC or locally. Prerequisites: generating and exporting imaging data and experimental metadata to ResearchDrive.
Useful links:
20211013-p04-NJW_932.yml). Each plate must have its own YAML file, even if parameters are identical across the batch.This approach allows you to run wrmXpress jobs on CHTC using just a folder of yml files on your local computer. It assumes (and checks) that the yml filenames match plate names with data and metadata stored on ResearchDrive. It uses Globus to transfer the relevant data and metadata files into the ResearchDrive CHTC folder, which allows direct transfers to your jobs using the Pelican platform, skipping the staging process.
Install the Globus CLI and log in: pip install globus-cli, then globus login. Run export PATH="$HOME/.local/bin:$PATH" if needed.
Setup persistent ssh authentication for CHTC.
Copy the relevant script files into a folder on your computer.
At the top of wrmx_submit.sh, set your CHTC NetID and access point. The Globus collection (wisc-drive-mzamanian) and the ResearchDrive path in wrmXpress.sub are already filled in.
Run chmod +x wrmx_submit.sh.
Run globus ls ab674f96-c05f-46af-a0cf-aa5fbabb65d1:/ and globus ls 798af6e9-d3bc-4c11-8b9d-a2a1eae74bb1:/ to check access to our ResearchDrive collections.
Put the yamls in a folder (e.g., ~/Desktop/ymls). Name each one <plate>.yml, matching the plate's folder name in ImageXpress/raw and ImageXpress/metadata. Then run: