ARBA00027624 organelle membrane (GO:0031090)

id: ARBA00027624
description: A highly complex rule with 304 condition sets that predicts GO:0031090 (organelle membrane)
  for proteins containing diverse domain families across 406 CATH FunFam domains, 68 InterPro domains,
  and spanning 63 taxonomic groups from genus to kingdom level.
status: COMPLETE
rule_type: ARBA
rule:
  rule_id: ARBA00027624
  condition_sets: []  # Too many to list (304 condition sets)
  go_annotations:
  - go_id: GO:0031090
    go_label: organelle membrane
    aspect: CC
  reviewed_protein_count: 0
  unreviewed_protein_count: 0
  created_date: '2021-10-20'
  modified_date: '2025-03-21'
  entries: []  # Cannot be computed due to rule complexity
review_summary: |
  ARBA00027624 is a severely problematic annotation rule that attempts to capture "organelle
  membrane" localization using an inappropriately complex design. The rule contains 304 condition
  sets spanning 406 CATH FunFam domains, 68 InterPro domains, 7 PANTHER families across 63
  taxonomic groups ranging from specific genera to entire kingdoms.

  The fundamental flaw lies in using GO:0031090 "organelle membrane" - an overly broad cellular
  component term that encompasses membranes from completely different organelles with distinct
  targeting mechanisms, import pathways, and biological functions. The rule conflates nuclear
  membrane proteins (using nuclear import machinery), mitochondrial membrane proteins (using
  TOM/TIM complexes), ER membrane proteins (using signal recognition particles), chloroplast
  proteins (using TOC/TIC complexes), and many others under a single annotation.

  Analysis reveals that 67% of condition sets (204/304) contain only single FunFam domains,
  indicating promiscuous annotation patterns likely to generate false positives. The taxonomic
  diversity is biologically unjustifiable, spanning plant-specific clades (BOP clade), mammalian
  orders (Primates, Rodentia), fungal classes (Saccharomycetes), and broad eukaryotic groups
  without mechanistic basis for such breadth.

  The rule violates multiple best practices: it exceeds complexity limits (>12 condition sets),
  uses inappropriate GO terms (too broad), lacks biological coherence, and creates high false
  positive risk. GO curators have raised concerns about such rules in the issue tracker, and
  this rule exemplifies why overly broad annotation approaches are problematic.

  Recommendation: Remove this rule entirely and replace with specific rules targeting individual
  organelles using appropriate GO terms and mechanistically coherent protein families.
action: DEPRECATE
action_rationale: |
  This rule represents a fundamental failure in annotation design with multiple critical issues:

  1. EXCESSIVE COMPLEXITY: 304 condition sets is 25 times larger than the recommended maximum,
     making the rule unmanageable and impossible to properly curate or validate.

  2. OVERLY BROAD GO TERM: GO:0031090 "organelle membrane" is far too general, encompassing
     membranes of nuclear envelope, mitochondria, ER, Golgi, peroxisomes, chloroplasts, and
     many others. This provides no meaningful biological insight.

  3. NO BIOLOGICAL COHERENCE: The rule groups together proteins solely based on potential
     membrane association without any shared biochemical function, cellular process,
     structural mechanism, or evolutionary relationship.

  4. MECHANISTIC INCOHERENCE: Different organelles use completely distinct targeting and
     insertion mechanisms (nuclear import, TOM/TIM complexes, signal recognition particles,
     TOC/TIC complexes, PTS targeting) that cannot be meaningfully grouped together.

  5. HIGH FALSE POSITIVE RISK: Many domains appear in cytoplasmic proteins, secreted proteins,
     or proteins with dual localizations that would be incorrectly annotated.

  6. INAPPROPRIATE TAXONOMIC SCOPE: The rule spans from individual genera (Mus, Arabidopsis)
     to entire kingdoms (Eukaryota) without biological justification for such diversity.

  This rule should be replaced with specific, mechanistically coherent rules targeting individual
  organelles with appropriate GO terms like GO:0005635 (nuclear envelope), GO:0005743
  (mitochondrial inner membrane), etc.
suggested_modifications:
- Deprecate the entire rule as it violates fundamental principles of GO annotation
- Replace with specific rules for individual organelles using appropriate GO terms
- Limit condition sets to <12 per rule with mechanistic coherence
- Use specific organelle membrane terms instead of the overly broad GO:0031090
parsimony:
  assessment: OVERLY_COMPLEX
  notes: |
    The rule contains 304 condition sets, which is 25 times larger than the recommended
    maximum of 12. This extreme complexity makes the rule unmanageable, impossible to
    properly validate, and computationally expensive to analyze. The large number of
    condition sets (406 unique CATH FunFam domains, 68 InterPro domains) suggests
    fundamental design problems rather than biological necessity.
literature_support:
  assessment: CONTRADICTED
  notes: |
    The rule contradicts established knowledge about organelle targeting mechanisms.
    Scientific literature clearly demonstrates that different organelles use distinct
    targeting and import mechanisms that cannot be grouped under a single annotation.
    GO curation guidelines recommend against using overly broad terms like GO:0031090,
    favoring specific organelle terms instead.
  supported_by:
  - reference_id: "file:rules/arba/ARBA00027624/ARBA00027624-deep-research-manual.md"
    supporting_text: "These are completely different molecular mechanisms that cannot be grouped under a single annotation rule"
condition_overlap:
  assessment: SIGNIFICANT
  notes: |
    With 406 unique CATH FunFam domains across 304 condition sets, there is likely
    substantial but unmeasurable overlap. The diversity of domains and the fact that
    67% of condition sets contain single domains suggests extensive redundancy and
    promiscuous annotation patterns.
  supported_by:
  - reference_id: "file:rules/arba/ARBA00027624/ARBA00027624-deep-research-manual.md"
    supporting_text: "204 condition sets (67%) contain only single FunFam domains, suggesting promiscuous annotation"
go_specificity:
  assessment: TOO_BROAD
  notes: |
    GO:0031090 "organelle membrane" is far too broad, encompassing membranes from nuclear
    envelope, mitochondria, ER, Golgi, peroxisomes, chloroplasts, and many other organelles.
    This term provides no meaningful biological insight and conflicts with GO guidelines
    that recommend specific subcellular localization terms.
  supported_by:
  - reference_id: "file:rules/arba/ARBA00027624/ARBA00027624-deep-research-manual.md"
    supporting_text: "GO:0031090 'organelle membrane' is an extremely broad cellular component term that encompasses membranes of virtually any membrane-bound organelle"
taxonomic_scope:
  assessment: TOO_BROAD
  notes: |
    The rule spans 63 taxonomic groups from individual genera (Mus, Arabidopsis) to
    entire kingdoms (Eukaryota) without biological justification. Different organelle
    targeting mechanisms evolved independently and are not conserved at such broad
    taxonomic levels. The scope conflates plant-specific organelles (chloroplasts)
    with animal-specific features without mechanistic basis.
  supported_by:
  - reference_id: "file:rules/arba/ARBA00027624/ARBA00027624-deep-research-manual.md"
    supporting_text: "The taxonomic scope spans: Broad eukaryotic clades (Eukaryota, Streptophyta), Specific mammalian orders (Primates, Rodentia), Plant-specific clades (BOP clade), Fungal classes (Saccharomycetes), Individual genera (Mus, Arabidopsis)"
confidence: 0.95
references:
- id: file:rules/arba/ARBA00027624/ARBA00027624-deep-research-manual.md
  title: Manual deep research analysis for ARBA00027624
  findings:
  - statement: Rule has 304 condition sets spanning 406 CATH FunFam domains, 68 InterPro domains, and 63 taxonomic groups, representing fundamental design flaws
  - statement: GO:0031090 is overly broad, encompassing membranes from nuclear envelope, mitochondria, ER, Golgi, peroxisomes, chloroplasts without mechanistic coherence
  - statement: Different organelles use completely distinct targeting mechanisms (nuclear import, TOM/TIM complexes, signal recognition particles, TOC/TIC complexes) that cannot be grouped together
supported_by:
- reference_id: file:rules/arba/ARBA00027624/ARBA00027624-deep-research-manual.md
  supporting_text: ARBA00027624 represents a fundamentally flawed annotation rule that violates GO annotation best practices