@prefix dcterms: <http://purl.org/dc/terms/> .
@prefix lokf: <https://w3id.org/lokf/> .
@prefix prov: <http://www.w3.org/ns/prov#> .
@prefix schema1: <http://schema.org/> .
@prefix skos: <http://www.w3.org/2004/02/skos/core#> .
@prefix xsd: <http://www.w3.org/2001/XMLSchema#> .

<https://turbomam.github.io/nmdc-lokf-demo/datasets/kbase-nmdc-arkin> a schema1:Dataset ;
    dcterms:requires <https://turbomam.github.io/nmdc-lokf-demo/glossary/berdl-tenant> ;
    schema1:about <https://turbomam.github.io/nmdc-lokf-demo/explanations/who-built-what-on-nmdc-data> ;
    schema1:creativeWorkStatus "draft" ;
    schema1:dateModified "2026-08-26T00:00:00+00:00"^^xsd:dateTime ;
    schema1:description "The Arkin Lab at LBNL has extended NMDC omics with vector embeddings, organism trait assignments, and a reconciled annotation layer." ;
    schema1:keywords "produced-by-arkin-lab" ;
    schema1:name "kbase.nmdc_arkin" ;
    schema1:text """# Contribution
The Arkin Lab at LBNL built this from NMDC omics, adding vector embeddings,
organism trait assignments and a reconciled annotation layer. It is hosted in the
`kbase` tenant, where NMDC-derived work lives.

Snapshot currency 2026-05-27 as recorded by the audit.""" ;
    prov:wasDerivedFrom <https://turbomam.github.io/nmdc-lokf-demo/datasets/nmdc-metadata> .

<https://turbomam.github.io/nmdc-lokf-demo/datasets/kbase-nmdc-neon> a schema1:Dataset ;
    dcterms:requires <https://turbomam.github.io/nmdc-lokf-demo/glossary/berdl-tenant> ;
    schema1:about <https://turbomam.github.io/nmdc-lokf-demo/explanations/who-built-what-on-nmdc-data> ;
    schema1:creativeWorkStatus "draft" ;
    schema1:dateModified "2026-08-26T00:00:00+00:00"^^xsd:dateTime ;
    schema1:description "Observations from the NSF National Ecological Observatory Network, processed through workflows of the kind NMDC uses and published alongside NMDC-derived products." ;
    schema1:keywords "produced-by-neon" ;
    schema1:name "kbase.nmdc_neon" ;
    schema1:text """# Contribution
NEON, an NSF program, collected these observations. They were processed through
workflows of the kind NMDC uses, which is what the `nmdc_` element of the name
records, and they are hosted in the `kbase` tenant alongside other derived work.

Three separate facts, all worth having: who collected the data, how it was
processed, and where it lives.""" .

<https://turbomam.github.io/nmdc-lokf-demo/datasets/nmdc-ncbi-biosamples> a schema1:Dataset ;
    dcterms:requires <https://turbomam.github.io/nmdc-lokf-demo/glossary/berdl-tenant> ;
    schema1:about <https://turbomam.github.io/nmdc-lokf-demo/explanations/who-built-what-on-nmdc-data> ;
    schema1:creativeWorkStatus "draft" ;
    schema1:dateModified "2026-08-26T00:00:00+00:00"^^xsd:dateTime ;
    schema1:description "NCBI BioSample records, hosted alongside NMDC records so the two can be queried together." ;
    schema1:keywords "produced-by-ncbi" ;
    schema1:name "nmdc.ncbi_biosamples" ;
    schema1:text """# Contribution
NCBI produces these records and remains their authority at
ncbi.nlm.nih.gov/biosample. They are hosted in the `nmdc` tenant so NMDC users can
query them beside NMDC's own records, which is a convenience for the reader rather
than a claim of authorship.""" .

<https://turbomam.github.io/nmdc-lokf-demo/datasets/nmdc-metadata> a schema1:Dataset ;
    dcterms:requires <https://turbomam.github.io/nmdc-lokf-demo/glossary/berdl-tenant> ;
    schema1:about <https://turbomam.github.io/nmdc-lokf-demo/explanations/who-built-what-on-nmdc-data> ;
    schema1:creativeWorkStatus "draft" ;
    schema1:dateModified "2026-08-26T00:00:00+00:00"^^xsd:dateTime ;
    schema1:description "Study, biosample, and omics records produced by NMDC, and the source kbase.nmdc_arkin is built from." ;
    schema1:keywords "produced-by-nmdc" ;
    schema1:name "nmdc.metadata" ;
    schema1:text """# Contribution
NMDC produces and curates these records, in the `nmdc` tenant it maintains. This
is the authoritative surface for NMDC data, and the starting point for the one derived
product this bundle records: the Arkin Lab's extended omics in the `kbase` tenant.""" .

<https://turbomam.github.io/nmdc-lokf-demo/explanations/who-built-what-on-nmdc-data> a lokf:Explanation ;
    schema1:creativeWorkStatus "draft" ;
    schema1:dateModified "2026-08-26T00:00:00+00:00"^^xsd:dateTime ;
    schema1:description "Work across DOE BER and partners sits alongside NMDC data in several different ways: derived from it, processed like it, hosted with it. Recording which is which keeps that ecosystem legible." ;
    schema1:genre "explanation" ;
    schema1:keywords "attribution",
        "berdl",
        "modelling",
        "reuse" ;
    schema1:name "Who built what on NMDC data" ;
    schema1:text """# Reuse is the point
NMDC publishes data so that other people can build on it. In the BERDL lakehouse the Arkin
Lab at LBNL has done exactly that, extending NMDC omics with vector embeddings, organism
trait assignments and a reconciled annotation layer.

Two other things sit nearby without being that. NEON observations are not NMDC data; they
are processed through workflows of the kind NMDC uses. NCBI BioSample records are not NMDC
data either; they are hosted alongside NMDC records so the two can be queried together.

Three different relationships, then: derived from, processed like, hosted with. All three are
what a shared lakehouse is for, and all three are worth having. But they are not the same
relationship, and a single phrase like "built on NMDC data" covers all three while
distinguishing none of them. This bundle exists to record which is which.

# Attribution needs somewhere to live
Contribution is a fact about a resource, and it has to be written down somewhere to be
usable. A database name is not that place, and was never meant to be: a name carries where
a resource is hosted, which is a different and also useful fact.

Both facts matter, and neither is derivable from the other. The Arkin Lab's derived
products are hosted under the `kbase` tenant, which is a deliberate arrangement: keeping
NMDC-derived work in `kbase` leaves the `nmdc` tenant a clean surface the NMDC team
curates. Hosting and contribution are two axes, and a single string cannot carry both
without losing one.

So the contributions go in the graph, as `prov:wasDerivedFrom` edges and producer
attributions, where they can be queried, cited and corrected. That is all this bundle does.""" .

<https://turbomam.github.io/nmdc-lokf-demo/glossary/berdl-tenant> a schema1:DefinedTerm ;
    schema1:creativeWorkStatus "draft" ;
    schema1:dateModified "2026-08-26T00:00:00+00:00"^^xsd:dateTime ;
    schema1:name "BERDL tenant" ;
    schema1:text """# Note
A tenant says where a database is hosted. It does not say who produced the data.""" ;
    skos:definition "The top-level namespace of a database in the BERDL lakehouse catalog, such as nmdc, kbase, globalusers, or a username." .

