Y24-368 - BGE blending labware creator

[BenTopping]

User story As a developer (Ben) I would like the new BGE pipeline to have a custom labware creator for the final blending step. This labware creator will take multiple pairs of pools: BGE Lib Pool (PCR Free MX pool) and a BGE Cap Lib Pool Norm (ISC MX pool) and create a BGE Blended cap lib pool.

This is a large story (and research) because of the unknowns and because it hasn't been done before in Limber. We will need to deal with tag clashes (maybe using tag_depth) because both pools will have the same exact tags (since they are from the same plate originally). There should be no active requests here but we might want to do an automated submission here for the final part of the pipeline, alternatively users can request additional sequencing for this final post-blending section.

Note if this is not possible we can change the end of the BGE pipeline to just be a normal ISC branch but blending is preferable if possible.

Who are the primary contacts for this story Matt M, Sara W, Liz C, Katy T, Ben T

Who is the nominated tester for UAT Matt M, Sara W, Liz C

Acceptance criteria To be considered successful the solution must allow:

• [ ] Using the finalised pipeline diagram for reference (linked below), add and test a labware creator for the blending step.
  • [ ] Pairs of pools to blend must be from the same initial plate (plate at which pipeline splits) and have matching tags. Tags may be missing from one side of the pool but that doesn't prevent blending.
  • [ ] There may be more than one pair of tubes depending on how many samples are being pooled together.

Dependencies This story is blocked by the following dependencies:

• Depends on Limber BGE pipeline story.

Additional context Research, question and answers, and lucid chart are available in this doc.

[KatyTaylor]

Bit more detail about the pairs of tubes.

• SSR will make a bulk submission for a batch of plates to go through the BGE pipeline.
• A bulk submission can actually contain multiple submissions - there's a column called 'submission name', which defines how many submissions are made on upload.
• SSRs use the submission name column to define which samples go into which pools. i.e. any rows with the same submission name will end up in the same 'BGE Lib Pool' and 'BGE Cap Lib Pool Norm' tubes.
• A batch of 5 full plates might end up with 2 pools/submissions per plate (on each branch), so 10 pairs of pool tubes in total.
• We don't currently have a way of linking the tubes from the same 'batch' together - there is no concept of batch here.
• So best we can do without changing that is probably to allow them to scan any tubes in & check the pairs match - but we won't be able to tell them if they've missed a pair of tubes from a batch, or if they've included a pair of tubes from the wrong batch. Not sure if that's OK or not.

[KatyTaylor]

There should be no active requests here but we might want to do an automated submission here for the final part of the pipeline, alternatively users can request additional sequencing for this final post-blending section.

See comment here - https://github.com/sanger/sequencescape/issues/4402#issuecomment-2475966858.