Many SmartSPIM mice have empty specimen_procedures

[dougollerenshaw]

There was a misunderstanding about the way that SmartSPIM specimen_procedures data was being stored in SLIMS when Jon and team built the feature in the metadata service to retrieve this data. @jtyoung84 said in a Teams message:

It was a misunderstanding on how the data is organized in Slims. We had assumed each ExperimentRun was associated with a single mouse brain, but it looks like there are multiple mouse ids.

Jon logged this as AllenNeuralDynamics/aind-slims-service#28 and fixed it in AllenNeuralDynamics/aind-slims-service#29.

For all affected mice, the SmartSPIM records will have an empty list in procedures.specimen_procedures. Two example subject_ids are 791054 and 777096.

Now that the metadata service has been patched to avoid this issue, it should be possible to create a migration script that would backfill procedures.specimen_procedures for all mice that are tracked in SLIMS. Note that this does not include smartspim mice that were imaged prior to the SLIMS tracking being in place.

Also note that there have been past migration scripts that have backfilled other procedures info such as viruses. We need to be ensure that this is not overwritten. Any migration scripts addressing this info should write to the procedures.specimen_procedures field only.

[dougollerenshaw]

Done! As noted in the PR, unfortunately most sessions have not been tracked in SLIMS, so even after this fix most are still missing specimen_procedures.