Class: TraitSources

Sources for trait data (databases, literature, predictions). Provenance information for trait assignments.

STATUS: PLANNED - Not yet implemented in NMDC database.

This table will track where trait annotations come from, distinguishing between curated assertions, rule-based predictions, and literature mining.

INTENDED SOURCE TYPES:

1. GOLD Curated Phenotypes (gold_curated)

_ - Source: GOLD database cvphenotype + organism_phenotype tables_

_ - Method: Manual curation by GOLD annotators_

_ - Coverage: 7,113 organism-phenotype mappings across 159 phenotype terms_

_ - Confidence: High (expert-curated)_

_ - Example: Picrophilus torridus -> Acidophile (curator-assigned)_

__

2. IMG Phenotype Rules (img_rules)

_ - Source: IMG phenotype_rule + phenotype_rule_taxons tables_

_ - Method: Pathway-based logical inference (65 rules)_

_ - Coverage: 2,059,203 taxon-trait predictions_

_ - Confidence: Medium-High (rule-based)_

_ - Rule example: "L-histidine auxotroph" = absence of pathway 162_

_ - Top predictions: L-histidine auxotroph (74,695 taxa),_

_ L-glutamate prototroph (73,791 taxa)_

__

3. Literature Mining (literature)

_ - Source: PubMed abstracts, full-text papers_

_ - Method: NLP extraction_

_ - Confidence: Variable (requires validation)_

__

4. Computational Prediction (computed)

_ - Source: Genome-based prediction tools_

_ - Method: ML models, homology-based inference_

_ - Confidence: Requires benchmarking_

__

IMG PHENOTYPE RULE EXAMPLES: | rule_id | name | rule | taxa_count | |---------|-------------------------------|--------------------------------|------------| | 22 | L-histidine auxotroph | !pathway(162) | 74,695 | | 2 | Aerobe | pathway(768|769|770) | 45,956 | | 4 | Denitrifier | pathway(764) AND (765) AND... | 1,653 | | 53 | Nitrogen fixer | pathway(798) | (variable) | | 52 | Dissimilatory sulfate reducer | pathway(870) AND (864) AND... | 626 | | 6 | Carbon fixation | pathway(527|335|285|...) | 483 |

URI: https://w3id.org/kbase/nmdc_core/TraitSources

classDiagram class TraitSources click TraitSources href "../TraitSources/" TraitSources : confidence_level TraitSources : coverage_count TraitSources : rule_definition TraitSources : source_database TraitSources : source_id TraitSources : source_name TraitSources : source_type

Slots

Name	Cardinality and Range	Description	Inheritance
source_id	1 String	Unique identifier for the trait source	direct
source_name	0..1 String	Human-readable source name	direct
source_type	0..1 String	Type of source (curated, rule_based, literature, computed)	direct
source_database	0..1 String	Database of origin	direct
rule_definition	0..1 String	For rule-based sources, the logical rule definition	direct
confidence_level	0..1 String	Confidence assessment of annotations from this source	direct
coverage_count	0..1 Integer	Number of trait-taxon assignments from this source	direct

Identifier and Mapping Information

Annotations

property	value
source_table	trait_sources
status	planned

Schema Source

from schema: https://w3id.org/kbase/nmdc_core

Mappings

Mapping Type	Mapped Value
self	https://w3id.org/kbase/nmdc_core/TraitSources
native	https://w3id.org/kbase/nmdc_core/TraitSources

LinkML Source

Direct

name: TraitSources
annotations:
  source_table:
    tag: source_table
    value: trait_sources
  status:
    tag: status
    value: planned
description: "Sources for trait data (databases, literature, predictions). Provenance\
  \ information for trait assignments.\nSTATUS: PLANNED - Not yet implemented in NMDC\
  \ database.\nThis table will track where trait annotations come from, distinguishing\
  \ between curated assertions, rule-based predictions, and literature mining.\nINTENDED\
  \ SOURCE TYPES:\n1. GOLD Curated Phenotypes (gold_curated)\n   - Source: GOLD database\
  \ cvphenotype + organism_phenotype tables\n   - Method: Manual curation by GOLD\
  \ annotators\n   - Coverage: 7,113 organism-phenotype mappings across 159 phenotype\
  \ terms\n   - Confidence: High (expert-curated)\n   - Example: Picrophilus torridus\
  \ -> Acidophile (curator-assigned)\n\n2. IMG Phenotype Rules (img_rules)\n   - Source:\
  \ IMG phenotype_rule + phenotype_rule_taxons tables\n   - Method: Pathway-based\
  \ logical inference (65 rules)\n   - Coverage: 2,059,203 taxon-trait predictions\n\
  \   - Confidence: Medium-High (rule-based)\n   - Rule example: \"L-histidine auxotroph\"\
  \ = absence of pathway 162\n   - Top predictions: L-histidine auxotroph (74,695\
  \ taxa),\n     L-glutamate prototroph (73,791 taxa)\n\n3. Literature Mining (literature)\n\
  \   - Source: PubMed abstracts, full-text papers\n   - Method: NLP extraction\n\
  \   - Confidence: Variable (requires validation)\n\n4. Computational Prediction\
  \ (computed)\n   - Source: Genome-based prediction tools\n   - Method: ML models,\
  \ homology-based inference\n   - Confidence: Requires benchmarking\n\nIMG PHENOTYPE\
  \ RULE EXAMPLES: | rule_id | name                          | rule              \
  \             | taxa_count | |---------|-------------------------------|--------------------------------|------------|\
  \ | 22      | L-histidine auxotroph         | !pathway(162)                  | 74,695\
  \     | | 2       | Aerobe                        | pathway(768|769|770)       \
  \    | 45,956     | | 4       | Denitrifier                   | pathway(764) AND\
  \ (765) AND...  | 1,653      | | 53      | Nitrogen fixer                | pathway(798)\
  \                   | (variable) | | 52      | Dissimilatory sulfate reducer | pathway(870)\
  \ AND (864) AND...  | 626        | | 6       | Carbon fixation               | pathway(527|335|285|...)\
  \       | 483        |"
from_schema: https://w3id.org/kbase/nmdc_core
attributes:
  source_id:
    name: source_id
    description: Unique identifier for the trait source
    examples:
    - value: gold_curated
      description: GOLD curator-assigned phenotypes
    - value: img_rule_2
      description: IMG aerobe prediction rule
    - value: img_rule_53
      description: IMG nitrogen fixer prediction rule
    - value: pubmed_nlp
      description: PubMed literature mining
    from_schema: https://w3id.org/kbase/nmdc_core
    rank: 1000
    identifier: true
    domain_of:
    - TraitSources
    - TraitTaxonomyMapping
    range: string
    required: true
  source_name:
    name: source_name
    description: Human-readable source name
    examples:
    - value: GOLD Curated Phenotypes
    - value: 'IMG Phenotype Rule: Aerobe'
    - value: 'IMG Phenotype Rule: Nitrogen Fixer'
    from_schema: https://w3id.org/kbase/nmdc_core
    rank: 1000
    domain_of:
    - TraitSources
    range: string
  source_type:
    name: source_type
    description: Type of source (curated, rule_based, literature, computed)
    examples:
    - value: curated
      description: Expert-curated assertions
    - value: rule_based
      description: Pathway-based logical rules
    - value: literature
      description: Literature-mined annotations
    - value: computed
      description: Computationally predicted
    from_schema: https://w3id.org/kbase/nmdc_core
    rank: 1000
    domain_of:
    - TraitSources
    range: string
  source_database:
    name: source_database
    description: Database of origin
    examples:
    - value: GOLD
    - value: IMG
    - value: PubMed
    from_schema: https://w3id.org/kbase/nmdc_core
    rank: 1000
    domain_of:
    - TraitSources
    range: string
  rule_definition:
    name: rule_definition
    description: 'For rule-based sources, the logical rule definition. Uses IMG pathway
      IDs with operators: | (OR), , (AND), ! (NOT).'
    examples:
    - value: (768|769|770)
      description: Aerobe = has cytochrome c oxidase pathway
    - value: (!162)
      description: His auxotroph = lacks histidine biosynthesis
    - value: (798)
      description: Nitrogen fixer = has nitrogenase pathway
    from_schema: https://w3id.org/kbase/nmdc_core
    rank: 1000
    domain_of:
    - TraitSources
    range: string
  confidence_level:
    name: confidence_level
    description: Confidence assessment of annotations from this source
    examples:
    - value: high
      description: Expert curated
    - value: medium-high
      description: Validated rule-based
    - value: medium
      description: Computational prediction
    from_schema: https://w3id.org/kbase/nmdc_core
    rank: 1000
    domain_of:
    - TraitSources
    range: string
  coverage_count:
    name: coverage_count
    description: Number of trait-taxon assignments from this source
    examples:
    - value: '7113'
      description: GOLD curated phenotypes
    - value: '2059203'
      description: IMG rule-based predictions
    from_schema: https://w3id.org/kbase/nmdc_core
    rank: 1000
    domain_of:
    - TraitSources
    range: integer

Induced

name: TraitSources
annotations:
  source_table:
    tag: source_table
    value: trait_sources
  status:
    tag: status
    value: planned
description: "Sources for trait data (databases, literature, predictions). Provenance\
  \ information for trait assignments.\nSTATUS: PLANNED - Not yet implemented in NMDC\
  \ database.\nThis table will track where trait annotations come from, distinguishing\
  \ between curated assertions, rule-based predictions, and literature mining.\nINTENDED\
  \ SOURCE TYPES:\n1. GOLD Curated Phenotypes (gold_curated)\n   - Source: GOLD database\
  \ cvphenotype + organism_phenotype tables\n   - Method: Manual curation by GOLD\
  \ annotators\n   - Coverage: 7,113 organism-phenotype mappings across 159 phenotype\
  \ terms\n   - Confidence: High (expert-curated)\n   - Example: Picrophilus torridus\
  \ -> Acidophile (curator-assigned)\n\n2. IMG Phenotype Rules (img_rules)\n   - Source:\
  \ IMG phenotype_rule + phenotype_rule_taxons tables\n   - Method: Pathway-based\
  \ logical inference (65 rules)\n   - Coverage: 2,059,203 taxon-trait predictions\n\
  \   - Confidence: Medium-High (rule-based)\n   - Rule example: \"L-histidine auxotroph\"\
  \ = absence of pathway 162\n   - Top predictions: L-histidine auxotroph (74,695\
  \ taxa),\n     L-glutamate prototroph (73,791 taxa)\n\n3. Literature Mining (literature)\n\
  \   - Source: PubMed abstracts, full-text papers\n   - Method: NLP extraction\n\
  \   - Confidence: Variable (requires validation)\n\n4. Computational Prediction\
  \ (computed)\n   - Source: Genome-based prediction tools\n   - Method: ML models,\
  \ homology-based inference\n   - Confidence: Requires benchmarking\n\nIMG PHENOTYPE\
  \ RULE EXAMPLES: | rule_id | name                          | rule              \
  \             | taxa_count | |---------|-------------------------------|--------------------------------|------------|\
  \ | 22      | L-histidine auxotroph         | !pathway(162)                  | 74,695\
  \     | | 2       | Aerobe                        | pathway(768|769|770)       \
  \    | 45,956     | | 4       | Denitrifier                   | pathway(764) AND\
  \ (765) AND...  | 1,653      | | 53      | Nitrogen fixer                | pathway(798)\
  \                   | (variable) | | 52      | Dissimilatory sulfate reducer | pathway(870)\
  \ AND (864) AND...  | 626        | | 6       | Carbon fixation               | pathway(527|335|285|...)\
  \       | 483        |"
from_schema: https://w3id.org/kbase/nmdc_core
attributes:
  source_id:
    name: source_id
    description: Unique identifier for the trait source
    examples:
    - value: gold_curated
      description: GOLD curator-assigned phenotypes
    - value: img_rule_2
      description: IMG aerobe prediction rule
    - value: img_rule_53
      description: IMG nitrogen fixer prediction rule
    - value: pubmed_nlp
      description: PubMed literature mining
    from_schema: https://w3id.org/kbase/nmdc_core
    rank: 1000
    identifier: true
    alias: source_id
    owner: TraitSources
    domain_of:
    - TraitSources
    - TraitTaxonomyMapping
    range: string
    required: true
  source_name:
    name: source_name
    description: Human-readable source name
    examples:
    - value: GOLD Curated Phenotypes
    - value: 'IMG Phenotype Rule: Aerobe'
    - value: 'IMG Phenotype Rule: Nitrogen Fixer'
    from_schema: https://w3id.org/kbase/nmdc_core
    rank: 1000
    alias: source_name
    owner: TraitSources
    domain_of:
    - TraitSources
    range: string
  source_type:
    name: source_type
    description: Type of source (curated, rule_based, literature, computed)
    examples:
    - value: curated
      description: Expert-curated assertions
    - value: rule_based
      description: Pathway-based logical rules
    - value: literature
      description: Literature-mined annotations
    - value: computed
      description: Computationally predicted
    from_schema: https://w3id.org/kbase/nmdc_core
    rank: 1000
    alias: source_type
    owner: TraitSources
    domain_of:
    - TraitSources
    range: string
  source_database:
    name: source_database
    description: Database of origin
    examples:
    - value: GOLD
    - value: IMG
    - value: PubMed
    from_schema: https://w3id.org/kbase/nmdc_core
    rank: 1000
    alias: source_database
    owner: TraitSources
    domain_of:
    - TraitSources
    range: string
  rule_definition:
    name: rule_definition
    description: 'For rule-based sources, the logical rule definition. Uses IMG pathway
      IDs with operators: | (OR), , (AND), ! (NOT).'
    examples:
    - value: (768|769|770)
      description: Aerobe = has cytochrome c oxidase pathway
    - value: (!162)
      description: His auxotroph = lacks histidine biosynthesis
    - value: (798)
      description: Nitrogen fixer = has nitrogenase pathway
    from_schema: https://w3id.org/kbase/nmdc_core
    rank: 1000
    alias: rule_definition
    owner: TraitSources
    domain_of:
    - TraitSources
    range: string
  confidence_level:
    name: confidence_level
    description: Confidence assessment of annotations from this source
    examples:
    - value: high
      description: Expert curated
    - value: medium-high
      description: Validated rule-based
    - value: medium
      description: Computational prediction
    from_schema: https://w3id.org/kbase/nmdc_core
    rank: 1000
    alias: confidence_level
    owner: TraitSources
    domain_of:
    - TraitSources
    range: string
  coverage_count:
    name: coverage_count
    description: Number of trait-taxon assignments from this source
    examples:
    - value: '7113'
      description: GOLD curated phenotypes
    - value: '2059203'
      description: IMG rule-based predictions
    from_schema: https://w3id.org/kbase/nmdc_core
    rank: 1000
    alias: coverage_count
    owner: TraitSources
    domain_of:
    - TraitSources
    range: integer