Quick start for developers

This page is for developers who need to produce ChemDCAT-AP-conformant output from a database or API. It assumes familiarity with data schemas (JSON Schema, YAML, JSON-LD) but not with LinkML or semantic web technologies.

LinkML in 60 seconds

ChemDCAT-AP is written in LinkML, a YAML-based schema language. If you know JSON Schema, the mental model transfers:

A class is like a JSON Schema object with typed properties. It defines a data shape.
A slot is a property on a class (a key in your JSON/YAML). It has a range (the expected type).
is_a is structural inheritance: a child class gets all the parent's slots.
A mixin injects slots into a class without being a parent. Think of it as a trait or interface.
class_uri and slot_uri are the ontology terms the class/slot maps to in RDF output. They do not affect the JSON/YAML structure.
inlined_as_list: true means the value is an array of nested objects (not references).
range is the type constraint: another class, a primitive (string, float, uriorcurie), or an enum.

For a full introduction, see the LinkML documentation.

CURIEs and prefixes

Throughout the schema and instance data you will see two forms of identifiers:

Full URIs: http://purl.obolibrary.org/obo/CHEBI_23367
CURIEs (Compact URI Expressions): CHEBI:23367

A CURIE is a shorthand. It consists of a prefix (CHEBI) and a local part (23367), separated by a colon. The prefix maps to a base URI declared in the schema's prefixes block. The CURIE expands by concatenating the base URI with the local part:

# In the schema:
prefixes:
  CHEBI: http://purl.obolibrary.org/obo/CHEBI_
  SIO: http://semanticscience.org/resource/SIO_
  qudt: http://qudt.org/schema/qudt/
  doi: https://doi.org/

# In instance data, these are equivalent:
id: CHEBI:23367           # → http://purl.obolibrary.org/obo/CHEBI_23367
id: SIO:000008            # → http://semanticscience.org/resource/SIO_000008
id: doi:10.14272/...      # → https://doi.org/10.14272/...

This matters for three reasons:

RDF output uses the expanded URIs. If you produce RDF directly (bypassing linkml-convert), you need to expand CURIEs yourself.
JSON-LD context is generated from the prefix map. If you serve JSON-LD, the @context handles CURIE expansion for consumers automatically.
Validation: linkml-validate understands CURIEs via the prefix map. Your YAML/JSON can use either form.

The prefix maps for ChemDCAT-AP are declared in each schema module (chem_dcat_ap.yaml, chemical_entities_ap.yaml, etc.) and include all ontology namespaces used in class_uri, slot_uri, and the broad, narrow or exact mappings.

What you produce, and how it becomes RDF

You write instance data as YAML or JSON conforming to the ChemDCAT-AP schema. The LinkML toolchain can then validate it, convert it to RDF (Turtle, JSON-LD, N-Triples). You do not need to write RDF directly.

Your DB/API → JSON or YAML → linkml-validate → linkml-convert → Turtle / JSON-LD

Alternatively, you can use the auto-generated JSON Schema to validate JSON output from your API directly, or the auto-generated SHACL shapes to validate RDF you produce by other means.

Accessing the schema and its artifacts

All schema artifacts (LinkML YAML schemas, JSON Schema, SHACL shapes, JSON-LD context, etc.) are accessible via content negotiation through the w3id PURLs:

DCAT-AP+: https://w3id.org/nfdi-de/dcat-ap-plus/
ChemDCAT-AP: https://w3id.org/nfdi-de/dcat-ap-plus/chemistry/

This is also how the schemas import each other.

Python dataclasses for programmatic instantiation are published on PyPI:

pip install dcat-ap-plus chem-dcat-ap

Three ways to use ChemDCAT-AP

1. Write YAML/JSON and validate with the CLI

Write instance data as YAML (as shown in the examples throughout this documentation), then validate. You must specify the target class (-C) that represents the root of your data:

uv run linkml validate tests/data/valid/SubstanceSampleCharacterizationDataset-001.yaml \
  -s src/chem_dcat_ap/schema/chem_dcat_ap.yaml \
  -C SubstanceCharacterizationDataset

QUDT vocabulary bindings

The schema declares bindings that constrain has_quantity_type and unit values to QUDT vocabularies. These are not yet enforced by linkml-validate. Validation via the linkml-term-validator is planned but not yet implemented in the ChemDCAT-AP pipeline. This means that, for now, values of these slots need to be validates externally.

Convert to RDF:

uv run linkml-convert -t ttl tests/data/valid/SubstanceSampleCharacterizationDataset-001.yaml \
  -s src/chem_dcat_ap/schema/chem_dcat_ap.yaml \
  -C SubstanceCharacterizationDataset
  -o SubstanceCharacterizationDataset-001.ttl

2. Use the Python dataclasses

import chem_dcat_ap
from chem_dcat_ap.datamodel.chem_dcat_ap import SubstanceSample, ChemicalEntity, MolecularFormula,
    InChIKey, Temperature
)

sample = SubstanceSample(
    id="https://example.org/sample-001",
    title="My sample",
    has_temperature=[
        Temperature(
            title= "my sample's temperature",
            value=300.0,
            has_quantity_type={"id": "http://qudt.org/vocab/quantitykind/Temperature"},
            unit={"id": "https://qudt.org/vocab/unit/K"},
        )
    ],
    composed_of=[
        ChemicalEntity(
            id="https://example.org/sample-001#compound",
            molecular_formula=[MolecularFormula(value="C11H12N2S")],
            inchikey=[InChIKey(value="UGRXAOUDHZOHPF-UHFFFAOYSA-N")],
        )
    ],
)

3. Restructure your existing JSON API

If you have an existing API that already serves chemical data in its own JSON format, you can restructure its output to conform to ChemDCAT-AP. Fetch the generated JSON Schema via content negotiation from the w3id PURL and use it as the target structure for your serializer. Once your output conforms, validate it with any JSON Schema validator.

If your pipeline produces RDF, use the generated SHACL shapes (also via content negotiation) with a processor like pyshacl to validate conformance.

The complete picture: Dataset to subject matter

The examples in the other documentation pages show individual classes (a MaterialSample, a SubstanceSample, a ChemicalReaction). Here is how they fit together in a complete Dataset structure using ChemDCAT-AP's convenience shapes. This is what your API ultimately needs to produce.

A reaction monitoring dataset

This is a real example from the Chemotion Repository (CRR-56408), showing a multi-step phosphine synthesis. The source file is part of ChemDCAT-AP's validation test suite: tests/data/valid/ReactionMonitoringDataset-001.yaml. The full example is extensive (~1400 lines) because it models individual reaction steps, work-up procedures, and purification as nested activities. The excerpt below shows the top-level Dataset structure and the first reaction step to illustrate the pattern.

# A ReactionMonitoringDataset - the Chemotion reaction DOI serves as the Dataset ID
id: doi:10.14272/reaction/SA-FUHFF-UHFFFADPSC-MTSVGCFANK-UHFFFADPSC-NUHFF-NUHFF-NUHFF-ZZZ
title:
  - "Dataset for Chemotion Repository synthesis CRR-56408"
description:
  - "This dataset contains the recorded information for the Chemotion Repository
     synthesis CRR-56408."
related_resource:
  - id: doi:10.1080/03086648208081193
    title: "SYNTHESE UND NMR-UNTERSUCHUNGEN VON 2-FLUOR-TRIPHENYLPHOSPHINEN..."
  - id: doi:10.1002/adsc.202400919
    title: "Arylation of Secondary Phosphines with Diaryliodonium Salts..."

# The full ChemicalReaction description lives here
is_about_activity:
  - id: https://www.chemotion-repository.net/pid/56408
    title:
      - "Chemotion Repository synthesis CRR-56408"
    description:
      - "The reaction has been conducted in dry glass ware under argon atmosphere..."
    rdf_type:
      id: RXNO:0000329
      title: "planned synthesis"
    # Participants, conditions, yield
    used_starting_material:
      - id: https://www.chemotion-repository.net/pid/56408_StartMat-1
        title: "StartMat-1: 1-bromo-2-fluorobenzene"
        has_mass:
          - title: "Mass [mg]"
            has_quantity_type: http://qudt.org/vocab/quantitykind/Mass
            unit: https://qudt.org/vocab/unit/MilliGM
            value: 1761.100
        composed_of:
          - id: https://pubchem.ncbi.nlm.nih.gov/compound/61259
            rdf_type:
              id: CHEBI:35496
              title: "fluorobenzenes"
    used_reactant: [...]
    used_solvent: [...]
    generated_product: [...]
    used_reactor: [...]
    has_yield:
      - value: 86
        has_quantity_type: http://qudt.org/vocab/quantitykind/Dimensionless
        unit: http://qudt.org/vocab/unit/PERCENT
    has_duration: PT22H
    # Reaction steps (chemical transformations)
    has_reaction_step:
      - id: https://www.chemotion-repository.net/pid/56408#step-1
        title: "Step 1: Lithiation and Phosphination"
        has_part:
          # step-1_part-1: dissolving of 1-bromo-2-fluorobenzene in anhydrous Tetrahydrofuran
          - id: https://www.chemotion-repository.net/pid/56408#step-1_part-1
            title:
              - "step-1_part-1: dissolving of 1-bromo-2-fluorobenzene in Tetrahydrofuran"
            rdf_type:
              id: CHMO:0002773
              title: "dissolving"
      - id: https://www.chemotion-repository.net/pid/56408#step-3
        title: "Step 3: Oxidation to Phosphine Oxide"
    # All steps including non-reaction activities (work-up, purification)
    has_part:
      - id: https://www.chemotion-repository.net/pid/56408#step-1
      - id: https://www.chemotion-repository.net/pid/56408#step-2
        rdf_type:
          id: OBI:0000094
          title: "material processing"
      - id: https://www.chemotion-repository.net/pid/56408#step-3
      - id: https://www.chemotion-repository.net/pid/56408#step-4
        rdf_type: 
          id: OBI:0000094
          title: "material processing"
      - id: https://www.chemotion-repository.net/pid/56408#step-5
        rdf_type: 
          id: CHMO:0002582
          title: "flash chromatography"

# The ReactionMonitoring activity -- references the reaction by IRI
was_generated_by:
  - id: doi:10.14272/reaction/SA-FUHFF-UHFFFADPSC-MTSVGCFANK-UHFFFADPSC-NUHFF-NUHFF-NUHFF-ZZZ#Monitoring
    title:
      - "Monitoring of the Chemotion Repo synthesis CRR-56408"
    evaluated_activity:
      - id: https://www.chemotion-repository.net/pid/56408

Key structural points:

This is a ReactionMonitoringDataset: was_generated_by points to a ReactionMonitoring and is_about_activity points to a ChemicalReaction.
The full reaction description (participants, conditions, yield, steps) lives on the is_about_activity node. The ReactionMonitoring activity under was_generated_by only references the reaction by IRI via evaluated_activity. This keeps one authoritative description in one place.
The ChemicalReaction uses two complementary nesting mechanisms: has_reaction_step for steps that are themselves chemical reactions (here lithiation and oxidation), and has_part for all steps including non-reaction activities (work-up, purification). Steps in has_reaction_step are also listed in has_part.
Non-reaction steps (warming, quenching, washing, drying, filtering, evaporation, chromatography) are modeled as generic Activity instances classified via rdf_type with OBI and CHMO terms.
Within each step, sub-steps (has_part on the step itself) model the fine-grained experimental procedure: dissolving, drop-wise addition, stirring, cooling, etc. Each sub-step is classified with an OBI or CHMO term (e.g., CHMO:0002773 dissolving, OBI:0000274 adding a material entity, CHMO:0002774 stirring) and declares its own inputs (had_input_entity) and outputs (had_output_entity), forming a complete input/output chain. Sub-steps link to their predecessor via had_input_activity, capturing the chronological sequence of the procedure.

Converting YAML to RDF

The reaction example is converted to Turtle using:

uv run linkml-convert -t ttl \
  tests/data/valid/ReactionMonitoringDataset-001.yaml \
  -s src/chem_dcat_ap/schema/chem_dcat_ap.yaml \
  -C ReactionMonitoringDataset

This produces the following Turtle (abbreviated for readability):

@base <https://search.nfdi4chem.de/dataset/> .
@prefix SIO: <http://semanticscience.org/resource/SIO_> .
@prefix RXNO: <http://purl.obolibrary.org/obo/RXNO_> .
@prefix RO: <http://purl.obolibrary.org/obo/RO_> .
@prefix BFO: <http://purl.obolibrary.org/obo/BFO_> .
@prefix dcat: <http://www.w3.org/ns/dcat#> .
@prefix dcterms: <http://purl.org/dc/terms/> .
@prefix prov: <http://www.w3.org/ns/prov#> .
@prefix schema: <http://schema.org/> .
@prefix xsd: <http://www.w3.org/2001/XMLSchema#> .

# The ReactionMonitoringDataset
<https://doi.org/10.14272/reaction/SA-FUHFF-UHFFFADPSC-MTSVGCFANK-...> a dcat:Dataset ;
    dcterms:title "Dataset for Chemotion Repository synthesis CRR-56408" ;
    dcterms:subject <https://www.chemotion-repository.net/pid/56408> ;
    theme:
        - preferred_label:
            - "Science and technology" ;
    prov:wasGeneratedBy <https://doi.org/10.14272/reaction/SA-FUHFF-UHFFFADPSC-MTSVGCFANK-UHFFFADPSC-NUHFF-NUHFF-NUHFF-ZZZ#Monitoring> .

# The ReactionMonitoring - references the reaction
<https://doi.org/10.14272/reaction/SA-FUHFF-UHFFFADPSC-MTSVGCFANK-UHFFFADPSC-NUHFF-NUHFF-NUHFF-ZZZ#Monitoring> a prov:Activity ;
    dcterms:title "Monitoring of the Chemotion Repo synthesis CRR-56408" ;
    prov:wasInformedBy <https://www.chemotion-repository.net/pid/56408> .

# The ChemicalReaction - full description
<https://www.chemotion-repository.net/pid/56408> a RXNO:0000329, SIO:010345 ;
    dcterms:title "Chemotion Repository synthesis CRR-56408" ;
    dcterms:description "The reaction has been conducted in dry glass ware under argon atmosphere. ..."
    schema:duration "PT22H"^^xsd:duration ;
    # Starting materials and reactants via 'has_starting_material' and 'has_reactant' which are both mapped to RO's 'has primary input'
    RO:0004009 <https://www.chemotion-repository.net/pid/56408_Reactant-1>,
        <https://www.chemotion-repository.net/pid/56408_Reactant-2>,
        <https://www.chemotion-repository.net/pid/56408_Reactant-3>,
        <https://www.chemotion-repository.net/pid/56408_StartMat-1>,
        <https://www.chemotion-repository.net/pid/56408_StartMat-2> ;
    # Product via 'has_product', which is mapped to RO's 'has primary output'
    RO:0004008 <https://dx.doi.org/10.14272/MTSVGCFANKMACF-UHFFFAOYSA-N.1> ;
    # AgenticEntites via DCAT-AP+ predicate
    prov:wasAssociatedWith <https://www.chemotion-repository.net/pid/56408_Reactor-1>,
        <https://www.chemotion-repository.net/pid/56408_Solvent-1>,
        <https://www.chemotion-repository.net/pid/56408_Solvent-2>,
        <https://www.chemotion-repository.net/pid/56408_Solvent-3>,
        <https://www.chemotion-repository.net/pid/56408_Solvent-4>,
        <https://www.chemotion-repository.net/pid/56408_Solvent-5>,
        <https://www.chemotion-repository.net/pid/56408_Solvent-6> .
    # Yield
    SIO:000008 [ a CHMO:0002855 ;
            dcterms:title "Yield [%]" ;
            dcterms:type VOC4CAT:0005005 ;
            qudt:hasQuantityKind <http://qudt.org/vocab/quantitykind/Dimensionless> ;
            qudt:unit <http://qudt.org/vocab/unit/PERCENT> ;
            prov:value "86.0"^^xsd:float ] ;
    # Reaction steps (chemical transformations) via has_reaction_step, which is mapped to BFO's 'has part' 
    BFO:0000051 <https://www.chemotion-repository.net/pid/56408#step-1>,
        <https://www.chemotion-repository.net/pid/56408#step-3> ;
    # All steps including work-up and purification via the 'has part' slot inherited from the Activity class
    dcterms:hasPart <https://www.chemotion-repository.net/pid/56408#step-1>,
        <https://www.chemotion-repository.net/pid/56408#step-2>,
        <https://www.chemotion-repository.net/pid/56408#step-3>,
        <https://www.chemotion-repository.net/pid/56408#step-4>,
        <https://www.chemotion-repository.net/pid/56408#step-5> ;

Notice how the YAML-to-RDF mapping works:

is_about_activity becomes dcterms:subject → the Dataset links to the ChemicalReaction.
was_generated_by becomes prov:wasGeneratedBy → the Dataset links to the ReactionMonitoring.
evaluated_activity becomes prov:wasInformedBy → the ReactionMonitoring references the ChemicalReaction.
has_reaction_step becomes BFO:0000051 (has part) → only chemical reaction steps.
has_part becomes dcterms:hasPart → all steps including non-reaction activities.
Reaction participants use domain-specific predicates: RO:0004009 for inputs, RO:0004008 for outputs, prov:wasAssociatedWith for agents.

A substance sample characterization dataset

This is a real example from the Chemotion Repository, showing an HSQC NMR measurement of a chemical substance sample. The source file is part of ChemDCAT-AP's validation test suite: tests/data/valid/SubstanceSampleCharacterizationDataset-001.yaml. Every CI run validates this file against the schema.

Note how the SubstanceSampleCharacterization activity uses the generic DCAT-AP+ attribute and agent slots (has_qualitative_attribute, has_quantitative_attribute, carried_out_by) rather than chemistry-specific sub-slots. This is deliberate: SubstanceSampleCharacterization is a coarse-grained convenience shape that leaves method-specific structuring to sub-profiles like NMR-DCAT-AP.

# A SubstanceSampleCharacterizationDataset
id: doi:10.14272/UGRXAOUDHZOHPF-UHFFFAOYSA-N/CHMO0000604
other_identifier:
  - notation: https://www.chemotion-repository.net/pid/50438
title:
  - "heteronuclear single quantum coherence (HSQC) dataset"
description:
  - "Dataset analysing Sample ID: CRS-50440 with heteronuclear single quantum
     coherence (HSQC) NMR spectroscopy"
theme:
  - preferred_label:
      - "Science and technology"

# Dual link: the Dataset is directly about the SubstanceSample
is_about_entity:
  - id: doi:10.14272/UGRXAOUDHZOHPF-UHFFFAOYSA-N.2
    title: "CRS-50440"
    description: "The analysed chemical substance sample CRS-50440."
    rdf_type:
      id: CHEBI:59999
      title: "chemical substance"
    composed_of:
      - id: doi:10.14272/UGRXAOUDHZOHPF-UHFFFAOYSA-N.2#EvaluatedCompound
        description: "compound assigned to doi:10.14272/UGRXAOUDHZOHPF-UHFFFAOYSA-N.2"
        other_identifier:
          - notation: https://pubchem.ncbi.nlm.nih.gov/compound/26248854
        inchikey:
          - value: "KVOIVNBYNQXCNY-BOCHJOTCSA-N"
            title: "assigned InChiKey"
        inchi:
            value: "InChI=1S/C11H12N2S/c1-12-7-10-8-14-11(13-10)9-5-3-2-4-6-9/..."
            title: "assigned InChi"
        smiles:
          - value: "CNCc1csc(n1)c1ccccc1"
            title: "assigned SMILES"
        molecular_formula:
          - value: "C11H12N2S"
            title: "assigned molecular formula"
        iupac_name:
          - value: "N-methyl-1-(2-phenyl-1,3-thiazol-4-yl)methanamine"
            description: Chemotion IUPAC name
          - value: "Methyl[(2-phenyl-1,3-thiazol-4-yl)methyl]amine"
            description: PubChem IUPAC name
        has_molar_mass:
          - has_quantity_type: http://qudt.org/vocab/quantitykind/MolarMass
            unit: https://qudt.org/vocab/unit/GM-PER-MOL
            value: 204.072119
            description: "Molar mass as specified in the Chemotion repository."
          - has_quantity_type: http://qudt.org/vocab/quantitykind/MolarMass
            unit: https://qudt.org/vocab/unit/GM-PER-MOL
            value: 204.29
            description: "Molar mass as specified in PubChem"

# The SubstanceSampleCharacterization activity that produced this dataset
was_generated_by:
  - id: doi:10.14272/UGRXAOUDHZOHPF-UHFFFAOYSA-N/CHMO0000604#X27932
    description:
      - "heteronuclear single quantum coherence (HSQC) NMR spectroscopy."
    title: X27932
    evaluated_entity:
      - id: doi:10.14272/UGRXAOUDHZOHPF-UHFFFAOYSA-N.2
        title: "CRS-50440"
    # Generic DCAT-AP+ attribute and agent slots are used here.
    # Sub-profiles like NMR-DCAT-AP should define more granular shapes.
    has_qualitative_attribute:
      - rdf_type:
          id: NMR:1400037
          title: "NMR pulse sequence"
        value: hsqcedetgp
        description: "used pulse program"
    has_quantitative_attribute:
      - rdf_type:
          id: NMR:1400025
          title: "sample temperature in magnet"
        description: "used sample temperature setting"
        has_quantity_type: http://qudt.org/vocab/quantitykind/Temperature
        unit: https://qudt.org/vocab/unit/K
        value: 300.1
      - rdf_type:
          id: NMR:1400087
          title: "number of scans"
        value: 2
        has_quantity_type: http://qudt.org/vocab/quantitykind/Count
        unit: http://qudt.org/vocab/unit/NUM
        description: "used number of scans"
    carried_out_by:
      - id: doi:10.14272/UGRXAOUDHZOHPF-UHFFFAOYSA-N/CHMO0000604#X27932_NMR_Spectrometer
        rdf_type:
          id: NMR:1000371
          title: "AVANCE III HD"
        title: "Avance III 400 NMR spectrometer"
        description: "used NMR spectrometer"
      - id: doi:10.14272/UGRXAOUDHZOHPF-UHFFFAOYSA-N/CHMO0000604#X27932_NMR_Solvent
        description: "used solvent"
        rdf_type:
          id: NCIT:C45790
          title: Solvent
        has_part:
          - id: https://pubchem.ncbi.nlm.nih.gov/compound/71583
            rdf_type:
              id: CHEBI:85365
              title: "deuterated chloroform"
            title: "chloroform-D1 (CDCl3)"

Key structural points:

This is a SubstanceSampleCharacterizationDataset: was_generated_by points to a SubstanceSampleCharacterization and is_about_entity points to a SubstanceSample.
The SubstanceSample uses rdf_type: CHEBI:59999 for additional typing beyond its schema-level class_uri: SIO:001378. This is the ClassifierMixin providing multi-typing.
The SubstanceSampleCharacterization activity uses generic DCAT-AP+ slots for its method-specific attributes (pulse program, temperature, scan count) and agents (spectrometer, solvent). This is intentional: the convenience shape is under-specified, leaving method-specific structuring to sub-profiles.
The evaluated_entity on the activity and is_about_entity on the dataset both point to the same SubstanceSample, implementing the dual linking pattern.

Converting YAML to RDF

The above YAML is converted to Turtle using linkml-convert:

uv run linkml-convert -t ttl \
  tests/data/valid/SubstanceSampleCharacterizationDataset-001.yaml \
  -s src/chem_dcat_ap/schema/chem_dcat_ap.yaml \
  -P "_base=https://search.nfdi4chem.de/dataset/" \
  -C SubstanceCharacterizationDataset

The flags:

-t ttl: output format is Turtle
-s: path to the ChemDCAT-AP schema
-P "_base=...": sets the RDF base URI (so relative id values resolve correctly)
-C SubstanceCharacterizationDataset: the target class (tells the converter which shape to convert against)

This produces the following Turtle (abbreviated for readability):

@base <https://search.nfdi4chem.de/dataset/> .
@prefix AFR: <http://purl.allotrope.org/ontologies/result#AFR_> .
@prefix BFO: <http://purl.obolibrary.org/obo/BFO_> .
@prefix SIO: <http://semanticscience.org/resource/SIO_> .
@prefix CHEBI: <http://purl.obolibrary.org/obo/CHEBI_> .
@prefix CHEMINF: <http://semanticscience.org/resource/CHEMINF_> .
@prefix CHMO: <http://purl.obolibrary.org/obo/CHMO_> .
@prefix NMR: <http://nmrML.org/nmrCV#NMR:> .
@prefix dcat: <http://www.w3.org/ns/dcat#> .
@prefix NCIT: <http://purl.obolibrary.org/obo/NCIT_> .
@prefix schema: <http://schema.org/> .
@prefix dcterms: <http://purl.org/dc/terms/> .
@prefix adms: <http://www.w3.org/ns/adms#> .
@prefix prov: <http://www.w3.org/ns/prov#> .
@prefix qudt: <http://qudt.org/schema/qudt/> .
@prefix xsd: <http://www.w3.org/2001/XMLSchema#> .
@prefix skos: <http://www.w3.org/2004/02/skos/core#> .

# The Dataset
<https://doi.org/10.14272/UGRXAOUDHZOHPF-UHFFFAOYSA-N/CHMO0000604> a dcat:Dataset ;
    dcterms:description "Dataset analysing Sample ID: CRS-50440 with heteronuclear single quantum coherence (HSQC) NMR spectroscopy" ;
    dcterms:subject <https://doi.org/10.14272/UGRXAOUDHZOHPF-UHFFFAOYSA-N.2> ; # Linking to the evaluated entity
    dcterms:title "heteronuclear single quantum coherence (HSQC) dataset" ;
    adms:identifier [ a adms:Identifier ;
            skos:notation "https://www.chemotion-repository.net/pid/50438" ] ;
    dcat:theme [ a skos:Concept ;
            skos:prefLabel "Science and technology" ] ;
    prov:wasGeneratedBy <https://doi.org/10.14272/UGRXAOUDHZOHPF-UHFFFAOYSA-N/CHMO0000604#X27932> .

# The SubstanceSample (dual-typed: SIO:001378 from class_uri + CHEBI:59999  from rdf_type)
<https://doi.org/10.14272/UGRXAOUDHZOHPF-UHFFFAOYSA-N.2> a CHEBI:59999, SIO:001378 ;
    dcterms:description "The analysed chemical substance sample CRS-50440." ;
    dcterms:title "CRS-50440" ;
    BFO:0000051 <https://doi.org/10.14272/UGRXAOUDHZOHPF-UHFFFAOYSA-N.2#EvaluatedCompound> .

# The ChemicalEntity with structure descriptors and quantitative attributes (all via SIO:000008)
<https://doi.org/10.14272/UGRXAOUDHZOHPF-UHFFFAOYSA-N.2#EvaluatedCompound> a CHEBI:23367 ;
    dcterms:description "compound assigned to doi:10.14272/UGRXAOUDHZOHPF-UHFFFAOYSA-N.2" ;
    SIO:000008 [ a CHEMINF:000059 ; prov:value "KVOIVNBYNQXCNY-BOCHJOTCSA-N" ],
               [ a CHEMINF:000018 ; prov:value "CNCc1csc(n1)c1ccccc1" ],
               [ a CHEMINF:000042 ; prov:value "C11H12N2S" ],
               [ a AFR:0002409 ;
                 dcterms:description "Molar mass as specified in PubChem" ;
                 qudt:hasQuantityKind <http://qudt.org/vocab/quantitykind/MolarMass> ;
                 prov:value "204.072119"^^xsd:float ] .

# The SubstanceSampleCharacterization activity (generic DCAT-AP+ predicates & dual-typed: CHMO:0000604 from rdf_type)
<https://doi.org/10.14272/UGRXAOUDHZOHPF-UHFFFAOYSA-N/CHMO0000604#X27932> a CHMO:0000604, prov:Activity ;
    dcterms:description "heteronuclear single quantum coherence (HSQC) NMR spectroscopy." ;
    dcterms:relation [ a NMR:1400087, qudt:Quantity ;  # dual-typed: NMR:1400087 from rdf_type
            dcterms:description "used number of scans" ;
            qudt:hasQuantityKind <http://qudt.org/vocab/quantitykind/Count> ;
            qudt:unit <http://qudt.org/vocab/unit/NUM> ;
            prov:value "2.0"^^xsd:float ],
        [ a NMR:1400037, prov:Entity ; # dual-typed: NMR:1400037 from rdf_type
            dcterms:description "used pulse program" ;
            prov:value "hsqcedetgp" ],
        [ a NMR:1400025, qudt:Quantity ; # dual-typed: NMR:1400025 from rdf_type
            dcterms:description "used sample temperature setting" ;
            qudt:hasQuantityKind <http://qudt.org/vocab/quantitykind/Temperature> ;
            qudt:unit <https://qudt.org/vocab/unit/K> ;
            prov:value "300.1"^^xsd:float ] ;
    dcterms:title "X27932" ;
    # Linking to the evaluated SubstanceSample 
    prov:used <https://doi.org/10.14272/UGRXAOUDHZOHPF-UHFFFAOYSA-N.2> ;
    # Linking to the agents involved 
    prov:wasAssociatedWith <https://doi.org/10.14272/UGRXAOUDHZOHPF-UHFFFAOYSA-N/CHMO0000604#X27932_NMR_Solvent>,
        <https://doi.org/10.14272/UGRXAOUDHZOHPF-UHFFFAOYSA-N/CHMO0000604#X27932_NMR_Spectrometer> .

<https://doi.org/10.14272/UGRXAOUDHZOHPF-UHFFFAOYSA-N/CHMO0000604#X27932_NMR_Solvent> a NCIT:C45790, prov:Agent ;
    dcterms:description "used solvent" ;
    dcterms:hasPart <https://pubchem.ncbi.nlm.nih.gov/compound/71583> .

<https://doi.org/10.14272/UGRXAOUDHZOHPF-UHFFFAOYSA-N/CHMO0000604#X27932_NMR_Spectrometer> a NMR:1000371, prov:Agent ;
    dcterms:description "used NMR spectrometer" ;
    dcterms:title "Avance III 400 NMR spectrometer" .

# `DefinedTerm` instances (ontology classes) used for additional typing
NMR:1000371 a schema:DefinedTerm ;
    schema:name "AVANCE III HD" .
NMR:1400025 a schema:DefinedTerm ;
    schema:name "sample temperature in magnet" .
NMR:1400037 a schema:DefinedTerm ;
    schema:name "NMR pulse sequence" .
NMR:1400087 a schema:DefinedTerm ;
    schema:name "number of scans" .
CHEBI:59999 a schema:DefinedTerm ;
    schema:name "chemical substance" .
CHEBI:85365 a schema:DefinedTerm ;
    schema:name "deuterated chloroform" .
CHMO:0000604 a schema:DefinedTerm ;
    schema:name "heteronuclear single quantum coherence" .
NCIT:C45790 a schema:DefinedTerm ;
    schema:name "Solvent" .

Notice how the YAML-to-RDF mapping works:

class_uri determines rdf:type: SubstanceSample becomes SIO:001378, ChemicalEntity becomes CHEBI:23367. The rdf_type slot adds additional types (e.g. CHEBI:59999).
slot_uri determines RDF predicates: composed_of becomes BFO:0000051, structure descriptors become SIO:000008, generic attributes become dcterms:relation, agents become prov:wasAssociatedWith.
DefinedTerm instances (from rdf_type, has_quantity_type, unit) become typed resources with schema:name for the title.
CURIEs expand via the prefix map: doi:10.14272/... becomes https://doi.org/10.14272/....

For the full documentation of these shapes and when to use them vs. defining your own, see Dataset and activity shapes.

Key types for serialization

When building a serializer, you need to know which fields expect which data types:

Field	Type	Example
`id`	URI or CURIE	`https://example.org/sample-001` or `doi:10.14272/...`
`value` on `QuantitativeAttribute`	`float`	`300.0`
`value` on `QualitativeAttribute`	`string`	`"UGRXAOUDHZOHPF-UHFFFAOYSA-N"`
`has_quantity_type`	`DefinedTerm` (object with `id` and optional `title` & `from_CV`)	`{"id": "http://qudt.org/vocab/quantitykind/Temperature"}`
`unit`	`DefinedTerm`	`{"id": "https://qudt.org/vocab/unit/K"}`
`rdf_type`	`DefinedTerm`	`{"id": "CHMO:0000595", "title": "..."}`
`type`	`DefinedTerm`	`{"id": "http://example.org/vocab/...", "title": "..."}`
`has_duration`	`xsd:duration` string	`"PT22H"`, `"P2D"`, `"PT1H30M"`
`has_physical_state`	enum value	`"SOLID"`, `"LIQUID"`, `"GASEOUS"`, `"PLASMA"`

DefinedTerm is a lightweight object for referencing ontology classes or vocabulary concepts. At minimum it needs an id (the URI or CURIE of the term). title is optional but recommended for human readability.

Where to find valid QUDT URIs

Every QuantitativeAttribute needs a has_quantity_type (what kind of quantity) and a unit (what unit it's measured in). Both are expected to be QUDT URIs (validation not yet implemented → see above warning):

Quantity kinds: browse at qudt.org/vocab/quantitykind/. Common ones: Temperature, Mass, Volume, Pressure, AmountOfSubstance, Density, MolarMass, Dimensionless.
Units: browse at qudt.org/vocab/unit/. Common ones: K, DEG_C, GM, MilliGM, L, MilliL, MOL, MilliMOL, BAR, GM-PER-MOL, PERCENT.

Development workflow and CI

The ChemDCAT-AP repository uses just as a command runner. The key recipes for developers:

Command	What it does
`just install`	Install project dependencies via `uv sync`
`just gen-project`	Generate all artifacts: JSON Schema, SHACL, OWL, Python dataclasses, Pydantic models
`just test`	Run the full test suite: schema generation, Python unit tests, and validation of all example files
`just testdoc`	Build and serve the documentation locally via MkDocs

The test pipeline (just test) validates every YAML file in tests/data/ against the schema and converts them to JSON and YAML output. Files in tests/data/invalid/ are expected to fail validation. This runs on every push to a PR, so only PRs that pass the full suite are merged.

If you want to try the examples locally:

git clone https://github.com/nfdi-de/chem-dcat-ap.git
cd chem-dcat-ap
just install
just test

Quick start for developers

LinkML in 60 seconds

CURIEs and prefixes

What you produce, and how it becomes RDF

Accessing the schema and its artifacts

Three ways to use ChemDCAT-AP

1. Write YAML/JSON and validate with the CLI

2. Use the Python dataclasses

3. Restructure your existing JSON API

The complete picture: Dataset to subject matter

A reaction monitoring dataset

Converting YAML to RDF

A substance sample characterization dataset

Converting YAML to RDF

Key types for serialization

Where to find valid QUDT URIs

Development workflow and CI

Further reading