Validate the plate-to-set metadata used for matched controls
Source:R/experiment_metadata.R
validate_metadata.RdOne row must represent one independent plate photographed once. set_id defines which treatment and control plates are biologically comparable; controls are never borrowed from a different set. Use globally unique plate IDs and unambiguous relative file paths. Repeated photographs of the same plate are not new replicates.
Usage
validate_metadata(
metadata,
files = NULL,
control = "control",
stage = c("comparison", "measurement")
)Arguments
- metadata
Data frame containing file, plate_id, set_id and treatment. treatment must equal control for controls. Other labels identify treatments consistently across sets. Optional zymo_strain, medium and day must be complete and constant within each set once any value is recorded; entirely unrecorded optional columns may remain blank.
- files
Optional vector of image paths. Exactly these images must appear once in metadata; no unmatched or unused rows are accepted.
- control
Exact label used for control plates, default control.
- stage
comparison (default) checks the matched experimental design. measurement checks image identities only; experimental labels and optional context fields may remain blank. Explicit invalid layouts are still errors.