ProteinTraitsMech — protein sequence & structure traits

Browse records →Corpus map →Protein map →Sequence map →

Record browser

Faceted, client-side browser over all protein-trait records — search by label, identifier, or definition; filter by axis, category, and source. Each record shows semantically related traits from a local text-embedding model.

Three 2-D maps sit alongside it: the corpus map projects a retained subset of record text embeddings (full-record or definition-only), the protein map lays out 68,657 Swiss-Prot proteins by the traits they carry, coloured by domain of life, and the sequence map lays out the canonical-example proteins by their ESM-2 protein-language-model embedding — sequence only, no annotation in the input — coloured by domain of life and filterable by sequence length and, for the half of them that carry one, by CATH class.

View schema →Example →

Schema & docs

The LinkML schema (ProteinTraitRecord, causal graphs), a worked P25888 example, and the curation model.

Explore SLC10 evidence →

Molecular explanations pilot

A research comparison of SLC10 proteins: experimentally observed contacts, derived residue correspondence, assay results, and source-backed mechanisms. Conserved sites are not automatically functional sites.

View on GitHub →

GitHub

Source YAML records, schema, build scripts, and the curation workflow.

Trait axes

SEQUENCE — motifs, signal/leader peptides, propeptides, cleavage sites, low-complexity / disordered regions, tandem repeats, compositional biases, conserved regions, epitopes, PTM sites.
STRUCTURE — folds, structural domains, secondary-structure arrangements, topology classes, quaternary state, subunit interfaces, active / binding / allosteric / metal sites, disulfide bonds, cavities, symmetry, dynamics, structural stability.
SEQUENCE_STRUCTURE — traits with both a sequence signature and 3D periodicity: coiled coils, and structural tandem repeats with demonstrated periodicity (RepeatsDB → MIXED_STRUCTURAL_REPEAT).
FUNCTION — enzymatic activity, binding capacity, cofactor requirement, subcellular localisation, environmental response, interaction partner. Grounded by EC / Rhea / ChEBI / GO / UniProt SubCell.
EVOLUTION — comparative-genomics traits: a protein's conservation across taxa (conserved, clade-specific, variable) and pangenome partition (core, soft-core, shell, cloud, persistent, singleton).

Trait categories

Every record carries a fine-grained trait_category. Counts link into the browser pre-filtered to that subset.

Current trait categories
Category	Axis	Records
`SEQ_DOMAIN`	SEQUENCE	75,935
`STRUCT_FOLD`	STRUCTURE	55,735
`SEQ_FAMILY`	SEQUENCE	36,096
`FUNC_PROTEIN_FAMILY`	FUNCTION	32,616
`FUNC_ORTHOLOG_GROUP`	FUNCTION	31,819
`FUNC_PATHWAY`	FUNCTION	28,098
`FUNC_ENZYMATIC_ACTIVITY`	FUNCTION	26,003
`STRUCT_INTERFACE`	STRUCTURE	20,639
`FUNC_INTERACTION_PARTNER`	FUNCTION	20,579
`SEQ_EPITOPE`	SEQUENCE	20,000
`STRUCT_HOMOLOGOUS_SUPERFAMILY`	STRUCTURE	15,177
`STRUCT_DOMAIN`	STRUCTURE	13,514
`FUNC_MOLECULAR_FUNCTION`	FUNCTION	10,041
`FUNC_RESISTANCE`	FUNCTION	7,452
`STRUCT_BINDING_SITE`	STRUCTURE	6,021
`SEQ_HOMOLOGOUS_SUPERFAMILY`	SEQUENCE	5,699
`STRUCT_TOPOLOGY`	STRUCTURE	5,427
`FUNC_LOCALIZATION`	FUNCTION	4,076
`SEQ_MOTIF`	SEQUENCE	3,121
`FUNC_TRANSPORT`	FUNCTION	2,285
`SEQ_REPEAT`	SEQUENCE	2,087
`SEQ_PTM_SITE`	SEQUENCE	1,252
`STRUCT_ACTIVE_SITE`	STRUCTURE	1,004
`SEQ_CLEAVAGE_SITE`	SEQUENCE	841
`SEQ_CONSERVATION`	SEQUENCE	775
`SEQ_MODIFIED_RESIDUE`	SEQUENCE	618
`STRUCT_CLASS`	STRUCTURE	299
`STRUCT_METAL_SITE`	STRUCTURE	292
`FUNC_COFACTOR_REQUIREMENT`	FUNCTION	287
`MIXED_TRANSMEMBRANE`	SEQUENCE_STRUCTURE	262
`SEQ_DISORDER`	SEQUENCE	202
`FUNC_INTERACTION_TYPE`	FUNCTION	147
`SEQ_ACTIVE_SITE`	SEQUENCE	133
`MIXED_STRUCTURAL_REPEAT`	SEQUENCE_STRUCTURE	122
`SEQ_GLYCOSYLATION_SITE`	SEQUENCE	85
`SEQ_BINDING_SITE`	SEQUENCE	82
`FUNC_BINDING_CAPACITY`	FUNCTION	69
`SEQ_CROSSLINK_SITE`	SEQUENCE	69
`STRUCT_ARCHITECTURE`	STRUCTURE	64
`FUNC_ENVIRONMENTAL_RESPONSE`	FUNCTION	55
`SEQ_LIPIDATION_SITE`	SEQUENCE	40
`STRUCT_STABILITY`	STRUCTURE	37
`STRUCT_SECONDARY`	STRUCTURE	34
`SEQ_TARGETING_SIGNAL`	SEQUENCE	31
`SEQ_LEADER_PEPTIDE`	SEQUENCE	20
`STRUCT_DYNAMICS`	STRUCTURE	18
`STRUCT_SURFACE`	STRUCTURE	16
`STRUCT_QUATERNARY`	STRUCTURE	11
`SEQ_COMPOSITION`	SEQUENCE	10
`STRUCT_SYMMETRY`	STRUCTURE	9
`MIXED_COILED_COIL`	SEQUENCE_STRUCTURE	7
`SEQ_TRANSIT_PEPTIDE`	SEQUENCE	7
`STRUCT_CAVITY`	STRUCTURE	7
`EVO_CONSERVATION`	EVOLUTION	6
`EVO_PANGENOME`	EVOLUTION	6
`SEQ_PROPEPTIDE`	SEQUENCE	6
`SEQ_SIGNAL_PEPTIDE`	SEQUENCE	6
`SEQ_INITIATOR_METHIONINE`	SEQUENCE	5
`SEQ_MATURE_CHAIN`	SEQUENCE	5
`STRUCT_ALLOSTERIC_SITE`	STRUCTURE	5
`STRUCT_DISULFIDE`	STRUCTURE	5
`SEQ_LOW_COMPLEXITY`	SEQUENCE	4
`SEQ_NONSTANDARD_RESIDUE`	SEQUENCE	4

Corpus at a glance — sources

Current source counts come from the same index as the record browser. Integration notes and directories retain source context; each count links to the matching source filter.

Current corpus sources
Source	Records	Integration notes and directories
LinkML LSF	29,202	LinkML LocalStructuralFeature valueset `data/traits/structure/{secondary,active_site,binding_site,cavity,disulfide,metal_site,dynamics,interface}/`
PROSITE	6,174	PROSITE PATTERN (generic) `data/traits/sequence/pattern/` PROSITE PATTERN routed to PTM subtypes `data/traits/sequence/{modified_residue,glycosylation,crosslink}/` PROSITE MATRIX (profile) `data/traits/sequence/profile/` PROSITE ProRule (DC=Domain) `data/traits/structure/domain/` PROSITE ProRule (DC=Site, keyword-routed) `data/traits/sequence/{modified_residue,glycosylation,prorule}/`
TED	13,860	TED novel folds (Zenodo v5) `data/traits/structure/fold/novel/` TED highly-symmetric folds `data/traits/structure/fold/high_symmetry/`
PSI-MOD	1,971	PSI-MOD (CC-BY-4.0) `data/traits/sequence/{modified_residue,glycosylation,lipidation,crosslink,ptm_ontology}/`
ECOD	45,113	ECOD v295 (A/X/H/T/F hierarchy) `data/traits/structure/{architecture,homologous_superfamily,topology,fold/ecod}/`
SCOPe	22,810	SCOPe 2.08 — Class/Fold/Superfamily/Family/Domain (Berkeley; instances px/sp excluded) `data/traits/structure/{class,fold,homologous_superfamily,domain}/scope/`
M-CSA	1,003	M-CSA (CC-BY-4.0) `data/traits/structure/active_site/mcsa/`
DisProt	35	DisProt — 32 IDPO disorder classes (structural state/transition/function; proteins as examples) (CC-BY-4.0) `data/traits/sequence/disorder/`
PSI-MI	147	PSI-MI interaction types (CC-BY-4.0) `data/traits/function/interaction_partner/psi_mi/`
METPO	119	METPO ecophysiological traits (growth preferences, tolerances, metabolism; CC-BY-4.0) `data/traits/function/{environmental_response,enzymatic_activity}/metpo/`
PATO	37	PATO physicochemical qualities (CC-BY-4.0) `data/traits/structure/{stability,dynamics,surface}/pato/`
curated	145	Curated stability taxonomy — per-condition (thermal, oxidative, saline, pH, osmotic, pressure, desiccation, chemical, proteolytic, mechanical) × increased/decreased (historical records retain CC0; current project contributions: CC BY 4.0) `data/traits/structure/stability/conditions/` Curated evolutionary / pangenome traits — conserved, clade-specific, variable; pangenome core/soft-core/shell/cloud/persistent/singleton (historical records retain CC0; current project contributions: CC BY 4.0) `data/traits/evolution/{conservation,pangenome}/` Curated RiPP leader-peptide classes — lanthipeptide/lasso/sactipeptide/… (historical records retain CC0; current project contributions: CC BY 4.0) `data/traits/sequence/leader_peptide/`
InterPro	26,264	InterPro entries — Domain / Homologous-superfamily / Repeat / Conserved-/Active-/Binding-site / PTM (public domain; GO via interpro2go → mapped_xrefs) `data/traits/{structure,sequence}/…/interpro/`
Pfam	31,025	Pfam-A families — Domain/Family→domain, Repeat, Coiled-coil, Disordered, Motif (public domain; GO+InterPro via pfam2go/pfam2interpro → mapped_xrefs) `data/traits/{structure/domain,sequence/repeat,mixed/coiled_coil,sequence/disorder,sequence/motif}/pfam/`
CATH	8,151	CATH-Gene3D hierarchy — Class/Architecture/Topology/Homologous-superfamily (CC-BY 4.0) `data/traits/structure/{class,architecture,topology,homologous_superfamily}/cath/`
Reactome	2,883	Reactome pathways — Homo sapiens reference set (CC0) `data/traits/function/pathway/reactome/`
CARD/ARO	7,452	CARD/ARO — antibiotic-resistance determinants + mechanisms (CC-BY 4.0) `data/traits/function/resistance/aro/`
TCDB	2,285	TCDB transport classification — Class/Subclass/Family (CC-BY-SA 3.0; 946 families ChEBI-grounded) `data/traits/function/transport/tcdb/`
COG	4,903	COG 2020 orthologous groups + 26 functional categories (US Gov public domain) `data/traits/function/ortholog_group/cog/`
Rhea	18,558	Rhea reactions — enzymatic reactions + ChEBI participants (CC-BY 4.0; EC via rhea2ec) `data/traits/function/enzymatic_activity/rhea/`
ExPASy ENZYME	7,375	ExPASy ENZYME — complete EC hierarchy (CC-BY 4.0; GO/RHEA mapped, KEGG direct, DR examples) `data/traits/function/enzymatic_activity/ec/`
RepeatsDB	122	RepeatsDB — structural tandem-repeat Class/Topology/Fold/Clan (CC-BY 4.0) `data/traits/sequence_structure/structural_repeat/repeatsdb/`
NCBIfam	38,394	NCBIfam (ex-TIGRFAMs) — prokaryotic family HMMs (US-gov public domain) `data/traits/structure/domain/ncbifam/`
CDD	38,218	CDD — NCBI-curated conserved domains + PRK clusters + KOG ortholog groups (US-gov public domain) `data/traits/{structure/domain,function/ortholog_group}/cdd/`
IDEAL	1	IDEAL — ProS (protean segment) trait; 1,448 IDPs as examples (CC-BY 4.0) `data/traits/sequence/disorder/`
ELM	353	ELM (Eukaryotic Linear Motif) — targeting/cleavage/motif classes (⚠ non-commercial) `data/traits/sequence/{targeting_signal,cleavage_site,ptm_site,motif}/elm/`
MEROPS	370	MEROPS peptidase families — S/C/A/M/… catalytic classes (EBI; academic) `data/traits/structure/domain/merops/`
HAMAP	2,394	—
PANTHER	15,489	—
PRINTS	2,106	—
SFLD	303	—
UniProtKB	5	—
eLife metallophores	20	—
manual	102,090	—
Total corpus	429,377

Retired integration: UniProtKB FT/CC/GO demultiplexer (seed_uniprot.py) — demo retired; real entries now attach as canonical_examples on class traits via fetch_uniprot_examples.py

Curation model

Every record is a stand-alone YAML with an identifier (source-anchored CURIE when possible), label + definition with provenance, trait_axis + trait_category, parent_traits / xrefs / synonyms, an optional sequence_pattern, optional canonical_examples (UniProtKB exemplars), optional evidence-backed causal_graphs (every edge cites ≥1 reference), and an append-only curation_history. Seeded records land as SEEDED; curator review flips them to REVIEWED.