Closed TWJW-SANGER closed 2 weeks ago
Maybe also fold in why they do purpose changes. Maybe fake cherrypick is for partials, purpose changes for full plates.
See Confluence page "Fake Cherry Picking (CP) Investigation"
Seems like we need to do 3 things (?):
Allow submissions from stock plates. Allow users to change a plate purpose and/or allow plates to proceed to any purpose on a specified pipeline. Allow plates to be submitted to SciOps with samples on for multiple studies.
Retention instructions on stock plates needs to be considered - if plates are used as stock plates but not called stock plates, will this cause confusion? (From conversation with Liz C)
Initially we could try adding Stock Plate to the list of acceptable purposes for limber PCR bespoke request type and also to the relationships for bespoke pcr as a starter for 10.
This should work for 10x - CITEseq and EMSeq.
The most effective solution for CASM would be to be able to submit samples under multiple studies. My initial investigation suggests that this would require major surgery for the following reasons:
On the plus side a request has a study which could be used to allow multiple studies for a submission.
As a rough estimate we would need to change the sample manifest and order models to support multiple studies. This would require data migrations. I presume that there would also be knock on effects including UI changes.
This is clearly an issue that is not going to go away so we would either need to consider this large amount of work or use an alternative solution.
It is also clear that submissions are complex by their nature.
Hi-C - Doesn't actually exist as a pipeline in Sequencescape or Limber so only has a library type. Not sure what we could even do in this case and it is only used for when TopUp sequencing is needed for some samples - rare occurence. This is a long read pipeline. There is another story for Hi-C that may cover this.
Outstanding issues:
As SSRs we would like to not have to perform fake cherry picks on stock plates to make submissions so that we are not recording fake data.
Description We support a number of pipelines whose first step requires a submission on the stock plates received in SciOps from faculty. A restriction within SequenceScape prevents users making submissions on stock plates - forcing SSRS to create fake transfers and plates, on which they can then create the desired submission.
This undesirable because:
Desired Outcome
Who the primary contacts are for this work SSR team PSD
Knowledge or Stake holders Liz C