Need
Not all metadata fields are applicable to all file types (e.g., reference_genome is relevant for BAM files but not for FASTQ). We need to review how this is currently handled in the Terra Data Repository (TDR) schema for the AnVIL project, specifically:
-
How not_applicable, not_available, and unspecified values are represented (if at all).
-
Whether there is a standardized mechanism for marking a field as not applicable.
If no consistent method exists, we should define a clear and interoperable approach for representing these cases across metadata fields and file types.