Segmentation Mask Metadata Attributes
The attributes shown here have been colleced for Segmentation Mask data, available from the HuBMAP Search and Entity APIs at Dataset.metadata.<attribute>.
These attributes are harmonized across all versions of the Segmentation Mask metadata specifications. See the latest version of the Segmentation Mask Metadata Specifications for the schema and directory structure needed when ingesting Segmentation Mask metadata.
* indicates a required field
| Attribute | Type | Description | Allowable Values |
|---|---|---|---|
| parent_dataset_id | The unique identifier from HuBMAP or SenNet for the dataset(s) used to generate this derived dataset. For example, in the case of a segmentation mask, the parent dataset would be the imaging assay containing the OME TIFF file utilized. If the derived dataset originates from multiple parent datasets, list them separated by commas. Example: HBM386.ZGKG.235 | ||
| lab_id | An identifier assigned by the data provider to reference an external metadata record for the dataset. The external record may be maintained independently and can support dataset traceability and provenance tracking. Leave this field empty if no such identifier exists. | ||
| derived_dataset_protocol_doi | The DOI for the protocols.io page that outlines the analysis conducted to generate this dataset. For instance, in the analysis of a sequencing dataset, the protocol should detail the pipeline executed, reference libraries utilized, and URLs for the tool programs. | ||
| derived_dataset_type | The specific type of analysis being conducted to generate the derived dataset. Example: Segmentation Mask | Object x Analyte Segmentation Mask |
|
| program_url | This is a link to the code or program run to generate this dataset (e.g., GitHub). This URL needs to contain the version of the program or code used to generate this derived dataset. For example on GitHub this should be to the tag or release for the actual version of the code within the respository that was used, rather than a more general link to the repository (e.g., https://github.com/hubmapconsortium/salmon-rnaseq/releases/tag/v2.1.15). If multiple programs are used to make this derived dataset then each URL should be included in double quotes, using a “;” (semicolon) as a delimiter between the quoted URLs. For example an entry might be included as follows: “https://www.github.com/blah/blah”; “https://www.github.com/foo/bar”. | ||
| contributors_path | The name of the file containing the ORCID IDs for all contributors to this dataset. Example: ./contributors.csv | ||
| data_path | The top-level directory containing the raw and/or processed data. For a single dataset upload, this might be represented as “.”, whereas for a data upload containing multiple datasets, this would be the directory name for the respective dataset. For example, if the data is within a directory named “TEST001-RK”, use the syntax “./TEST001-RK” for this field. If there are multiple directory levels, use the format “./TEST001-RK/Run1/Pass2”, where “Pass2” is the subdirectory where the single dataset’s data is stored. This is an internal metadata field used solely for data ingestion. Example: ./TEST001-RK | ||
| is_cell_type_annotation_included | Indicates whether cell type annotations are included in the dataset. Answer with “Yes” or “No” value. Example: Yes | Yes No |
|
| metadata_schema_id | The unique string identifier for the metadata specification version, which is easily interpretable by computers for purposes of data validation and processing. Example: 22bc762a-5020-419d-b170-24253ed9e8d9 |