Single-cell Schema

Contents

Single-cell Schema#

The cell object acts as point of reference for all data that can be related to an individual cell, either by direct observation or inference.

File Format Specification#

Files are YAML/JSON with an AIRR Data File structure. Files should be encoded as UTF-8. Identifiers are case-sensitive. Files should have the extension .yaml, .yml, or .json.

Schema Field Definitions#

Cell Fields#

Download as TSV

Name	Type	Attributes	Definition
`cell_id`	string	required, identifier	Identifier for the Cell object. This identifier must be unique within a given study, but it is recommended that it be a universally unique record locator to enable database applications.
`repertoire_id`	string	required, nullable	Identifier to the associated repertoire in study metadata.
`data_processing_id`	string	optional, nullable	Identifier of the data processing object in the repertoire metadata for this cell.
`receptors`	array of string	optional, nullable	Array of receptor identifiers defined for the Receptor objects associated with this cell
`cell_subset`	Ontology	optional, nullable	Commonly-used designation of isolated cell population.
`cell_phenotype`	string	optional, nullable	List of cellular markers and their expression levels used to isolate the cell population.
`cell_label`	string	optional, nullable	Free text cell type annotation. Primarily used for annotating cell types that are not provided in the Cell Ontology.
`virtual_pairing`	boolean	required, nullable	boolean to indicate if pairing was inferred.
`cell_type`	string	optional, nullable	Cell type (source). Most often the type should be “observed,” meaning it corresponds to an actual physical sample that was sequenced. Other allowed values are “simulated,” i.e. the Cells were generated in silico and there is no linked Subject/Sample metadata, and “inferred” for Cells that are phylogenetically reconstructed from observed sequences. Inferred Cells should be assigned to a Repertoire with repertoire_type=inferred.

Expression Fields#

Download as TSV

Name	Type	Attributes	Definition
`expression_id`	string	required, identifier	Identifier for the Expression object. This identifier must be unique within a given study, but it is recommended that it be a universally unique record locator to enable database applications.
`cell_id`	string	required	Identifier of the cell to which this expression data is related.
`repertoire_id`	string	required, nullable	Identifier for the associated repertoire in study metadata.
`data_processing_id`	string	required, nullable	Identifier of the data processing object in the repertoire metadata for this cell.
`property_type`	string	required	Keyword describing the property type and detection method used to measure the property value. The following keywords are recommended, but custom property types are also valid: “mrna_expression_by_read_count”, “protein_expression_by_fluorescence_intensity”, “antigen_bait_binding_by_fluorescence_intensity”, “protein_expression_by_dna_barcode_count” and “antigen_bait_binding_by_dna_barcode_count”.
`property`	Ontology	required, nullable	Name of the property observed, typically a gene or antibody identifier (and label) from a canonical resource such as Ensembl (e.g. ENSG00000275747, IGHV3-79) or Antibody Registry (ABREG:1236456, Purified anti-mouse/rat/human CD27 antibody).
`value`	number	required, nullable	Level at which the property was observed in the experiment (non-normalized).