New special interest group: ontologies

[mirpedrol]

I confirm my proposal adheres to the nf-core special interest group requirements and recommendations

• I confirm

Special interest group name

ontologies

What is it about?

I would like to propose a special interest group about ontologies.
We already have the channel #ontologies, which could be renamed.

The group is aimed at discussing everything related to ontologies and nf-core modules.

Working objectives:

• Tagging the operation of nf-core modules
• Tagging the input and output of the modules
• Developing tools to tag nf-core modules with ontologies (current development of edamMCP)
• Developing tools to contribute back to the EDAM database (or other ontologies)

Topics of discussion:

• Exploring how ontologies can help with pipeline chaining
• Exploring and discussing how ontologies interact with Nextflow Typing
• Best structure of the meta.yml files

There is an ongoing collaboration effor with Biocondactor, EDAM and bio.tools communities and we are holding bi-monthly meetings. It would be good to have a group of nf-core people interested on these topics to ensure that the efforts also benefit our community.

Keywords

ontology, modules

Why do we need a new special interest group?

There has been some work on tagging nf-core modules with EDAM ontology terms, and during the nf-core hackathon we had a group working on the edamontology/edammcp tool. It would be good to have a place to keep track of the progress and allow people to continue their contribution to this.

Who would be interested?

• People who work with ontologies
• People who want to use tagged nf-core modules eg. for module grouping, for module chaining...

Who would lead?

@mirpedrol

Pre-existing participating consortia, if relevant

there is an existing #ontologies channel in nf-core and an ongoing collaboration with the bioconductor and EDAM communities

List of typical nf-core pipelines used

none

[JoseEspinosa]

Should the scope be limited to modules? Maybe ontologies could be used at some point at the pipeline and/or subworkflow level couldn't they?

[JoseEspinosa]

In any case, I
/approve

[nf-core-bot]

Approval status: ✅ Approved

SIG proposal requires approvals from 2 core team members.

Current approvals: 3/2

Review Status	Core Team members
✅ Approved	@JoseEspinosa, @mashehu, @jfy133
🕐 Pending	@ewels, @maxulysse, @mribeirodantas, @mirpedrol, @FriederikeHanssen, @kenibrewer, @edmundmiller, @sateeshperi, @FranBonath, @christopher-hakkaart, @nvnieuwk

[ewels]

Is this more of a working group, versus a special interest group? I feel like it's a project rather than a shared interest.

[mirpedrol]

A working group fits too, it's definitely more like a project, but it can also be extended to other sub-projects

[mirpedrol]

I’ve been giving this a bit more thought, and I think it fits better as a special interest group. It feels similar in scope to the regulatory group.

Right now, the focus is mainly on the edamMCP work that came out of the nf-core hackathon, but the interest could easily broaden. This could include tagging modules (my initial motivation), exploring pipeline chaining, collaborating more closely with EDAM or other ontology registries, and so on.

I’m currently collaborating with a group of people from Bioconductor, EDAM, and bio.tools, and we already hold bi-monthly meetings. My goal here is to create a more formal way to involve nf-core contributors in these ongoing efforts and in any new projects that might start.

Happy to hear others' thoughts :)

[mirpedrol]

more opinions @nf-core/core?

[mashehu]

/approve
if there are already meetings etc. we can use the website to advertise these.

[jfy133]

Based on the scope described here:

The group is aimed at discussing everything related to ontologies and nf-core modules.

    Tagging the operation of nf-core modules
    Tagging the input and output of the modules
    Developing tools to tag nf-core modules with ontologies
    Developing tools to contribute back to the EDAM database (or other ontologies)

I agree with @ewels this seems too focused on action orientated steps, thus a working group would be more fitting than a special interestr group. To me the latter should 'theoretically' involve broader discussions/coordination/a bit of theory etc. rather than building stuff.

I’m currently collaborating with a group of people from Bioconductor, EDAM, and bio.tools, and we already hold bi-monthly meetings. My goal here is to create a more formal way to involve nf-core contributors in these ongoing efforts and in any new projects that might start.

This sounds more Interest Group-ey though. So I'm inclined to give an approval based on that, but could you please provide an updated description?

[jfy133]

For @mirpedrol that is :)