The current QC documentation omits how the QC "Stage" should be used, and when we expect QC to be generated. This should be updated:
Raw: immediate QC that was generated alongside the rig or triggered by upload to S3
Processing: QC generated by generic pipelines that are run for this kind of data asset, e.g. spike sorting, motion correction, etc
Raw/Processing should only ever apply to data from one modality in a single data asset
Analysis: QC generated for a specific project's analysis scripts
Multi-session: Analysis QC that depend on comparisons with other data assets, e.g. checking that all FOVs are aligned or that a stitched chronic recording has matching units
The current QC documentation omits how the QC "Stage" should be used, and when we expect QC to be generated. This should be updated:
Raw/Processing should only ever apply to data from one modality in a single data asset