3388 allow organisms without consensus sequences

[fengelniederhammer]

resolves #3388

preview URL: https://3388-allow-organisms-with.loculus.org/

Summary

This allows having no (consensus) sequence data in sequence entries. Changes include:

• The respective organism needs to be configured with allowSubmissionOfConsensusSequences: false in the values.yml.
• The reference genomes can be empty ({ "nucleotideSequences": [], "genes": [] })
• The backend allows uploads without a sequence file for such organisms
  • The sequences will always be an empty object/map.
  • It doesn't validate that the preprocessing pipeline doesn't submit sequences.
• The website doesn't require a sequence file in uploads (and has slightly adaptes texts there)
  
• The edit page doesn't show sequence data
  
• The mutation filter is not shown
  
• The sequence details page doesn't show sequence data:
  
• The download dialog doesn't offer sequence downloads:
  

Screenshot

PR Checklist

• All necessary documentation has been adapted.
• The implemented feature is covered by an appropriate test.

[fengelniederhammer]

I just noticed that the backend still requires the submissionId in the metadata uploads. Actually they are not relevant, because the metadata entries don't need to be matched on sequences. But maybe it's still good to keep them (to have less complex changes and to have them still around when we e.g. allow uploading of raw sequences later)?

[chaoran-chen]

I agree, let's leave the submissionId mandatory. When we start uploading raw sequences, the ID will also be needed.

[fhennig]

You mentioned in the summary that the backend doesn't check that the pipeline doesn't upload sequences; does that mean that uploading sequences via API is still possible? What happens if sequences are still uploaded? Will the data just be in SILO but just generally ignored?

[fengelniederhammer]

You mentioned in the summary that the backend doesn't check that the pipeline doesn't upload sequences; does that mean that uploading sequences via API is still possible? What happens if sequences are still uploaded? Will the data just be in SILO but just generally ignored?

I would need to check. Maybe it's not even true what I wrote and the backend actually checks that there are no superfluous sequences. At the latest, SILO would complain.

[fhennig]

I still have to go through the kt files, but I wanted to submit a first batch of comments already.

[fhennig]

Thanks for making the changes! It looks good to me now. Also the Kotlin code looks good!

[fhennig]

still looks good to me 👍