NeST KG
The NeST (Nested Systems in Tumors) Knowledge Graph converts the NeST hierarchical map of cancer protein systems into RDF, together with the IAS (Integrated Association Stringency) protein-association network from which the map was derived. It captures the nested containment hierarchy of protein systems, their member proteins, per-system cancer-mutation statistics and MONDO-typed cancer associations, and ~210k scored protein-protein associations as reified statements.
The NeST (Nested Systems in Tumors) Knowledge Graph (NeST KG) is a semantic knowledge graph derived from the NeST hierarchical map of human cancer protein systems. NeST organizes proteins into a nested hierarchy of physical assemblies and functional systems, inferred from integration of large-scale protein-protein interaction and cancer proteomics data.
The knowledge graph is constructed by converting the NeST networks from the Cytoscape CX2 format into RDF (Resource Description Framework) using standard biological ontologies. The source networks are publicly available on NDEx (Network Data Exchange).
Key Features
- Nested System Hierarchy: Captures the containment relationships between protein systems, representing how smaller assemblies nest within larger functional systems
- System Membership: Associates proteins (identified by UniProt accessions, or numeric HGNC identifiers for the 12 symbols with no protein product) with the systems they belong to
- System Annotations: Every system carries its size (
ndexv:memberCount) and HiSig significance statistics (lasso weight, BH-adjusted p-value); systems significantly mutated in a cohort additionally carry cancer associations with a tumors-mutated fraction - Standard Ontologies: Uses Biolink Model predicates for containment (
biolink:part_of), system membership (biolink:has_member) and protein association (biolink:interacts_with), together with Biolink'sAssociationpattern for system–cancer links; Semanticscience Integrated Ontology (SIO) for entity typing; and MONDO and ECO for cancer types and evidence codes. - Querying:
biolink:interacts_withis symmetric but stored in one direction only — match both (?a ^biolink:interacts_with|biolink:interacts_with ?b). Per-edge IAS scores live on reifiedrdf:Statementnodes, not on the asserted triple.
Ontologies Used
| Prefix | Ontology | Usage |
|---|---|---|
| biolink | Biolink Model | Containment (part_of), system membership (has_member), protein associations (interacts_with), and association structure (Association, genetically_associated_with, knowledge_level, agent_type, has_evidence) |
| SIO | Semanticscience Integrated Ontology | Entity types — SIO:010043 (protein), SIO:010035 (gene) |
| UniProt | UniProt | Protein entity identifiers |
| HGNC | HGNC | Gene entity identifiers — the 12 symbols with no protein product |
| MONDO | Mondo Disease Ontology | Cancer types in system→disease associations (14 types, incl. pan-cancer MONDO:0004992) |
| ECO | Evidence & Conclusion Ontology | Evidence code on each association |
| STATO, OBI, NCIT | — | Referenced via rdfs:seeAlso on the statistical properties |
| ndexv, nestv, iasv | Minted (NDEx) | ndexv:ProteinSystem, ndexv:memberCount, HiSig weight/p-value, IAS score |
Funding
This work is supported by the NSF Proto-OKN (Prototype Open Knowledge Network) program, the NIH NCI NDEx project (5U24CA269436), and the NIH NHGRI Cytoscape project (5U24HG012107).
References
Zheng F, Kelly MR, Ramms DJ, Heintschel ML, Tao K, Tutuncuoglu B, Lee JJ, Ono K, Foussard H, Chen M, Herrington KA, Silva E, Liu SN, Chen J, Churas C, Wilson N, Kratz A, Pillich RT, Patel DN, Park J, Kuenzi B, Yu MK, Licon K, Pratt D, Kreisberg JF, Kim M, Swaney DL, Nan X, Fraley SI, Gutkind JS, Krogan NJ, Ideker T. Interpretation of cancer mutations using a multiscale map of protein systems. Science. 2021 Oct;374(6563):eabf3067. doi: 10.1126/science.abf3067. PMID: 34591613; PMCID: PMC8590742.
Pratt D, Chen J, Welker D, Rivas R, Pillich R, Rynkov V, Ono K, Miello C, Hicks L, Szalma S, Stojmirovic A, Dobrin R, Braxenthaler M, Kuentzer J, Demchak B, Ideker T. NDEx, the Network Data Exchange. Cell Syst. 2015 Oct 28;1(4):302-305. doi: 10.1016/j.cels.2015.10.001. PMID: 26594663; PMCID: PMC4649937.
Shannon P, Markiel A, Ozier O, Baliga NS, Wang JT, Ramage D, Amin N, Schwikowski B, Ideker T. Cytoscape: a software environment for integrated models of biomolecular interaction networks. Genome Res. 2003 Nov;13(11):2498-504. doi: 10.1101/gr.1239303. PMID: 14597658; PMCID: PMC403769.
Ono K, Fong D, Gao C, Churas C, Pillich R, Lenkiewicz J, Pratt D, Pico AR, Hanspers K, Xin Y, Morris J, Kucera M, Franz M, Lopes C, Bader G, Ideker T, Chen J. Cytoscape Web: bringing network biology to the browser. Nucleic Acids Res. 2025 Jul 7;53(W1):W203-W212. doi: 10.1093/nar/gkaf365. PMID: 40308211; PMCID: PMC12230733.
NeST systems significantly enriched for somatic mutation (HiSig adjusted p < 0.05)
PREFIX ndexv: <https://www.ndexbio.org/vocab/>
PREFIX nestv: <https://www.ndexbio.org/vocab/nest/>
PREFIX rdfs: <http://www.w3.org/2000/01/rdf-schema#>
SELECT ?system ?label ?adjP ?weight ?members WHERE {
?system a ndexv:ProteinSystem ;
rdfs:label ?label ;
ndexv:memberCount ?members ;
nestv:hisigAdjustedPValue ?adjP .
OPTIONAL { ?system nestv:hisigWeight ?weight }
FILTER(?adjP < 0.05)
}
ORDER BY ?adjP
LIMIT 100
graph TD
classDef projected fill:lightgreen;
classDef literal fill:orange;
classDef iri fill:yellow;
v1("?adjP"):::projected
v3("?label"):::projected
v4("?members"):::projected
v2("?system"):::projected
v5("?weight"):::projected
c3([https://www.ndexbio.org/vocab/ProteinSystem]):::iri
f0[["?adjP < '0.05^^xsd:decimal'"]]
f0 --> v1
v2 --"a"--> c3
v2 --"rdfs:label"--> v3
v2 --"https://www.ndexbio.org/vocab/memberCount"--> v4
v2 --"https://www.ndexbio.org/vocab/nest/hisigAdjustedPValue"--> v1
subgraph optional0["(optional)"]
style optional0 fill:#bbf,stroke-dasharray: 5 5;
v2 -."https://www.ndexbio.org/vocab/nest/hisigWeight".-> v5
end
Direct child subsystems of a given NeST system (one level down the hierarchy)
PREFIX ndexv: <https://www.ndexbio.org/vocab/>
PREFIX biolink: <https://w3id.org/biolink/vocab/>
PREFIX rdfs: <http://www.w3.org/2000/01/rdf-schema#>
PREFIX nest: <https://www.ndexbio.org/identifiers/>
SELECT ?child ?childLabel ?members WHERE {
?child biolink:part_of nest:NEST-1 ;
rdfs:label ?childLabel ;
ndexv:memberCount ?members .
}
ORDER BY DESC(?members)
LIMIT 100
graph TD
classDef projected fill:lightgreen;
classDef literal fill:orange;
classDef iri fill:yellow;
v2("?child"):::projected
v3("?childLabel"):::projected
v1("?members"):::projected
c2([https://www.ndexbio.org/identifiers/NEST-1]):::iri
v2 --"biolink:part_of"--> c2
v2 --"rdfs:label"--> v3
v2 --"https://www.ndexbio.org/vocab/memberCount"--> v1
All NeST systems a given protein belongs to, with mutation-enrichment status, keyed on UniProt
PREFIX ndexv: <https://www.ndexbio.org/vocab/>
PREFIX nestv: <https://www.ndexbio.org/vocab/nest/>
PREFIX biolink: <https://w3id.org/biolink/vocab/>
PREFIX rdfs: <http://www.w3.org/2000/01/rdf-schema#>
PREFIX uniprot: <http://purl.uniprot.org/uniprot/>
# Example protein: EGFR = uniprot:P00533. Swap in any UniProt accession.
SELECT ?system ?label ?adjP (IF(?adjP < 0.05, "enriched", "not enriched") AS ?mutationStatus) WHERE {
?system biolink:has_member uniprot:P00533 ;
rdfs:label ?label ;
nestv:hisigAdjustedPValue ?adjP .
}
ORDER BY ?adjP
LIMIT 100
graph TD
classDef projected fill:lightgreen;
classDef literal fill:orange;
classDef iri fill:yellow;
v1("?adjP"):::projected
v3("?label"):::projected
v4("?mutationStatus")
v2("?system"):::projected
c2([uniprotkb:P00533]):::iri
v2 --"biolink:has_member"--> c2
v2 --"rdfs:label"--> v3
v2 --"https://www.ndexbio.org/vocab/nest/hisigAdjustedPValue"--> v1
bind0[/"if(?adjP < '0.05^^xsd:decimal','enriched','not enriched')"/]
v1 --o bind0
bind0 --as--o v4
| SPARQL Endpoint | https://apps.okn.us/nestkg/sparql |
|---|---|
| Triple Pattern Fragments | https://apps.okn.us/ldf/nestkg |
| Class | Entities |
|---|
| Property | Triples |
|---|