[Programming question]: Galaxy workflows

[haesleinhuepf]

Do you know what Galaxy workflows (.ga files) are? If I outline a rough image processing workflow, could you construct a galaxy workflow from this?

git-bob think about this

[github-actions]

^{This message was generated by git-bob (version: 0.14.1dev, model: claude-3-5-sonnet-20241022, log), an experimental AI-based assistant. It can make mistakes and has limitations. Check its messages carefully.}

Hi! 👋

I know Galaxy workflows are JSON-formatted workflow files used with the Galaxy platform - a web-based platform for computational biomedical research. However, I'm not sufficiently familiar with their specific structure and requirements to reliably construct a .ga workflow file.

@haesleinhuepf - Could you help with guidance on Galaxy workflow construction?

Best,
git-bob

[haesleinhuepf]

Here is an example: https://github.com/rmassei/imaging_KNIME_to_Galaxy/blob/main/src/test_data/01_nuclei_segmentation/workflows/2024_nuclei_segmentation_galaxy.ga

Let's construct a new galaxy workflow with just two steps:

• Gaussian blur filter
• Threshold

git-bob think about this

[github-actions]

^{This message was generated by git-bob (version: 0.14.1dev, model: claude-3-5-sonnet-20241022, log), an experimental AI-based assistant. It can make mistakes and has limitations. Check its messages carefully.}

Based on the example, here's a Galaxy workflow file implementing a simple Gaussian blur and thresholding pipeline:

{
    "a_galaxy_workflow": "true",
    "format-version": "0.1",
    "name": "gaussian_blur_threshold",
    "steps": {
        "0": {
            "id": 0,
            "type": "data_input",
            "inputs": [{"name": "Input Image"}],
            "position": {"left": 0, "top": 0}
        },
        "1": {
            "id": 1,
            "tool_id": "toolshed.g2.bx.psu.edu/repos/imgteam/2d_simple_filter/ip_filter_standard/1.12.0+galaxy1",
            "type": "tool",
            "input_connections": {
                "input": {"id": 0, "output_name": "output"}
            },
            "tool_state": "{\"filter\": {\"filter_type\": \"gaussian\", \"size\": \"3.0\"}}",
            "position": {"left": 300, "top": 0}
        },
        "2": {
            "id": 2, 
            "tool_id": "toolshed.g2.bx.psu.edu/repos/imgteam/2d_auto_threshold/ip_threshold/0.18.1+galaxy3",
            "type": "tool",
            "input_connections": {
                "input": {"id": 1, "output_name": "output"}
            },
            "tool_state": "{\"th_method\": {\"method_id\": \"otsu\"}}",
            "position": {"left": 600, "top": 0}
        }
    }
}

[haesleinhuepf]

git-bob comment on ways to automatically check if this workflow is formally correct. Is there a python lib or CLI tool which does that?

[github-actions]

^{This message was generated by git-bob (version: 0.14.1dev, model: claude-3-5-sonnet-20241022, log), an experimental AI-based assistant. It can make mistakes and has limitations. Check its messages carefully.}

For checking if a Galaxy workflow file is formally correct, there are a couple of options:

1. gxformat2 Python package: It has tools for validating Galaxy workflow formats

from gxformat2.yaml import validate_workflow
validate_workflow("workflow.ga")

2. Galaxy's command line interface planemo:

planemo lint workflow.ga

Both tools can help catch syntax errors and validate the workflow structure.

Do you want me to demonstrate how to use any of these validation tools in more detail?

[haesleinhuepf]

Ok, let's write a Jupyter notebook that saves the worklow above in a .ga file and afterwards uses gxformat2 to validate the workflow.

git-bob implement this!

[haesleinhuepf]

Ok, this didn't work.

Let's try to use planemo instead. The correct syntax to use it for validating a workflow seems:

planemo workflow_lint path/to/workflow.ga

git-bob implement another notebook which writes the galaxy workflow above in a .ga file and validates it, this time using planemo CLI instead of gxformat2.