Development subsites

Creating a new subsite (a.k.a Lab)

Subsite (lab) landing pages are built and served from the Galaxy Media Site under /landing/<lab-name>. GMS has a framework for building new labs pages in a consistent and reproducible manner, with a schema written in Python.

Step-by-step

1. Create a git branch for the new lab.
2. Decide on a consisteny lab name to use. This will be referred to as $LAB here.
3. Create a new subsite dir by cloning webapp/home/subdomains/genome/ to webapp/home/subdomains/${LAB}.
4. Create webpage content for the new lab by modifying the ${LAB}/sections.py ${LAB}/content/*.py files, observing the Python schema (see below).
5. Clone the HTML template webapp/home/templates/home/subdomains/genome.html to ${LAB}.html.
6. Edit the HTML template to make the landing page for your lab. It will render the content created in the step above as sections/accordion elements.
7. You may wish to reuse/create snippets for this template in webapp/home/templates/home/subdomains/components.
8. Run the branch with local development server.
9. In the web admin, add a Subsite record with the lab name $LAB.
10. You should find the lab landing page at the URL /landing/$LAB.

The content schema

Content for the lab pages is defined in a Python schema (ideally this would be YAML, but since YAML can't handle variables natively this would lead to large/repetitive files.)

A basic understanding of Python is required to contribute to this content, but can be done by a non-developer (especially if they have Git familiarity). Ideally this will be the domain expert responsible for designing the page.

The Python schema goes something like this:

# As defined in sections.py:
sections = [
    {
        "id": str,
        "title": str,  # e.g. "Data import and preparation"
        "tabs": [
            {
                "id": str,            # e.g. "tools"
                "title": str,         # e.g. "Tools"
                "heading_md": str,  # accepts inline HTML e.g. <code> <b> <i> <a>
                "content": content,
            },
            # ... Each of the above will make a new tab in this section
        ],
    },
    # ... Each of the above will make a new section
]

# As defined in content/*.py:
content = [
    {
        "title_md": str,        # inline HTML
        "description_md": str,  # inline HTML; should be a <p>; can have internal line breaks etc.
        "inputs": [
            {
                'datatypes': [str, ],  # List of accepted Galaxy datatypes for this input
                'label': str,  # Name of this input
            },
            {   # An example input:
                'datatypes': ['fastq', 'bam', 'sam'],  # List of accepted datatypes for this input
                'label': 'Raw sequencing reads',  # Descriptive name for this input
            },
            # ...
        ],
        # Optional fields:
        "button_link": str,  # URL for run button e.g. Galaxy tool link
        "button_md": str,  # defaults to "play" icon
        "button_tip": str,   # tooltip text; defaults to "Run tool"
        "view_link": str,    # URL for view button e.g. workflowhub link
        "view_md": str,    # defaults to "eye" icon
        "view_tip": str,     # tooltip text; defaults to None
    },
    # ... Each of the above will make an accordion item
]

Some sections (e.g. "Help") may benefit from having a longer description with more inline HTML, for example:

    {
        "title_md": 'Can I use Apollo to  share and edit the annotated genome?',
        "description_md": """
            <p>
                Apollo is web-browser accessible system that lets you conduct real-time collaborative curation and editing of genome annotations.
              </p>
              <p>
                The Australian BioCommons and our partners at QCIF and Pawsey provide a hosted
                <a href="https://apollo-portal.genome.edu.au/" target="_blank">
                  Apollo Portal service
                </a>
                where your genome assembly and supporting evidence files can be hosted. All system administration is taken care of, so you and your team can focus on the annotation curation itself.
            </p>

            <p>
                This
                <a href="https://training.galaxyproject.org/training-material/topics/genome-annotation/tutorials/apollo-euk/tutorial.html" target="_blank">
                  Galaxy tutorial
                </a>
                provides a complete walkthrough of the process of refining eukaryotic genome annotations with Apollo.
            </p>
        """,
        "button_link": "https://support.biocommons.org.au/support/solutions/articles/6000244843-apollo-for-collaborative-curation-and-editing",
        "button_md": "More info",
    }

The button_link or view_link can be omitted to hide that button. The above schema will render a webpage that looks something like this:

For full examples of the above, please refer to genome/content/assembly.py.