Crowd curation

[chaoran-chen]

We often talked about crowd curation but haven't discussed much how we envision it in detail. This is to collect ideas of features that we might want. I started this list based on our very old requirements collection and extended it further.

As an authenticated user, I want to…

• propose changes to metadata (ex: date, location)
• propose changes to sequence data
• propose flagging a sequence
• add a comment about the proposal
• curate an entry from the website
• use external software (incl. self-written scripts) for curation (i.e. load sequences into the external software and submit curation results from the software) – in other words, I want to be able to curate through an API
• see proposals by other users
• support/up-vote curation changes proposed by others
• have a web page that displays my curation contribution (e.g., number of curated sequences, identified issues, etc.)

As a data submitter, I want to…

• see proposals to my sequences
• accept or reject a proposal
• get an email notification if there are new proposals for my sequences

As a maintainer, I want to…

• accept or disable the crowd curation feature for the whole instance or only some organisms

[anna-parker]

Thanks for starting this! I wanted to add some more points on the interaction with INSDC

Curation and the INSDC

As a curator

I want my curations to also appear in INSDC.

As a data owner

I want to see that a curator has proposed changes to my sequence in Pathoplexus even if I only submitted to INSDC and be able to accept/reject these changes.

As INSDC

Only ENA accepts curations of sequences in INSDC that have not been curated by a sequence owner - these are visible as an overlay.

As a maintainer (ingest)

I want to know which ingested metadata fields have been curated and which have been revised by a sequence owner.

As ENA solves this issue with keeping curations in an overlay and not as a separate version it would make sense to also do that for Loculus.

Proposal

Store curations in a separate DB table as a change log, where each curation is an overlay for a specific version and will be used by the subsequent versions of a sequence unless there are metadata field overlays in which case manual review will be required. In the mean time we can add a note on the website to inform users of the debate. (Note in our current state ingest will be forced to do that and create a changelog for each sequence that has been curated which is very tedious)

TO DO

• investigate curation submission in Consider ways of submitting curations to ENA Clearing House #3324