Improve specification of how WES outputs are structured

[ianfore]

Cross-referencing to ga4gh/cloud-interop-testing#111

[uniqueg]

The RO-Crate, WorkflowHub and ELIXIR Cloud communities are starting work on how WES runs can be described and packaged as RO-Crates, both prospective (i.e., runs to be executed) and retrospective (i.e., runs that have been executed). That work may be relevant to this discussion and anyone interested is very welcome to contribute. For now at least, work will be summarized here. There is also a dedicated Slack channel here. We will start with what we can get from current WES specs, but will probably propose some changes/additions for metadata that we may find are crucial but missing. In those cases, we will of course open relevant issues on here, and we will also keep an eye on relevant discussions in existing issues, Cloud and FASP calls etc.

Apart from WES champions, this may also be interesting for @denis-yuen, @briandoconnor and @dglazer

[stain]

WorkflowHub use the Workflow RO-Crate to capture a workflow of any type https://w3id.org/workflowhub/workflow-ro-crate/

RO-Crate have now defined https://www.researchobject.org/workflow-run-crate/profiles/ profiles for workflow run provenance

The TRE-FX project has additionally created a workflow run request crate profile: https://w3id.org/trusted-wfrun-crate/ used with WfExS - this also supports validation and verification phases.