SPKW Anopheles gambiae (ANOGA) Subproject

Parent project: SPKW.md

Overview

This subproject explores UniProt Keyword (SPKW) unique contributions to Anopheles gambiae (African malaria mosquito) GO annotations. ANOGA is of high interest because:

1. Vector biology relevance: Mosquito immune system and salivary proteins are critical for understanding malaria transmission
2. Less curated organism: Fewer experimental annotations means higher reliance on automated pipelines
3. Unique biology: Insect immunity, blood-feeding adaptations, and vector-pathogen interactions

Summary Statistics (from ANOGA.ddb)

Distribution by Aspect

Status

• [x] Initial exploration complete (2026-01-30)
• [x] D7 protein family review complete (2026-01-30) - ALL 8 PROTEINS REVIEWED
• [x] Immune gene pilot batch complete (2026-01-30) - 6 GENES REVIEWED
• [x] PGRP/TEP extended batch (2026-01-31) - ALL 8 GENES REVIEWED
• [ ] Review remaining immune genes (~133 more)

High-Interest Categories for ANOGA

These terms are uniquely relevant to mosquito biology:

Known Problematic Terms (from Human Analysis)

Query Used

-- Database: ~/repos/go-db/db/ANOGA.ddb
-- Find SPKW-unique contributions (gene-term pairs with no other evidence)

WITH spkw_annotations AS (
    SELECT a.db_object_id, a.db_object_symbol, a.ontology_class_ref, a.aspect
    FROM gaf_association a
    WHERE a.supporting_references = 'GO_REF:0000043'
),
gene_term_with_other_evidence AS (
    SELECT DISTINCT db_object_id, ontology_class_ref
    FROM gaf_association
    WHERE supporting_references != 'GO_REF:0000043'
),
spkw_unique AS (
    SELECT s.*
    FROM spkw_annotations s
    WHERE NOT EXISTS (
        SELECT 1 FROM gene_term_with_other_evidence o
        WHERE o.db_object_id = s.db_object_id
          AND o.ontology_class_ref = s.ontology_class_ref
    )
)
SELECT
    su.ontology_class_ref,
    tl.label,
    su.aspect,
    COUNT(DISTINCT su.db_object_id) as gene_count
FROM spkw_unique su
JOIN term_label tl ON su.ontology_class_ref = tl.id
GROUP BY su.ontology_class_ref, tl.label, su.aspect
ORDER BY gene_count DESC;

D7 Family Review Summary (8 proteins)

Result: 4/4 (100%) "toxin activity" annotations are OVER-ANNOTATED; the other 4 D7 proteins do not carry GO:0090729 in GOA

Why "Toxin Activity" is Wrong for D7 Proteins

The GO definition of toxin activity (GO:0090729) requires:

"initiating pathogenesis (leading to an abnormal, generally detrimental state)"

D7 proteins fail this criterion because:
1. They do NOT initiate pathogenesis - they facilitate normal feeding behavior
2. They do NOT cause tissue damage - effects are reversible
3. The host returns to normal after the mosquito finishes feeding
4. Their mechanism is sequestration (removing mediators), not introducing harmful substances

The Kratagonism Mechanism

D7 proteins are "kratagonists" (from Greek kratos = power/strength + antagonist) - high-affinity scavengers that neutralize host defense molecules:

Host response to bite:
  Tissue damage → Serotonin/Histamine release → Vasoconstriction + Inflammation
                                    ↓
                           D7 protein binds mediator
                                    ↓
                           Mediator sequestered (inactive)
                                    ↓
                           Blood vessel stays dilated → Successful feeding

This is fundamentally different from toxin activity:
- Toxin: Introduces harmful substance → causes pathology
- Kratagonist: Removes host substance → prevents host defense

Short-Form vs Long-Form D7 Proteins

Exception: D7L3 (7DL3) binds serotonin, making it unique among long-forms.

Recommended GO Annotations for D7 Proteins

Instead of "toxin activity":
- GO:0051378 (serotonin binding) - for D7r1-4, D7L3
- GO:0051381 (histamine binding) - for D7r1-4
- GO:1900047 (negative regulation of hemostasis) - for all
- GO:0005615 (extracellular space) - appropriate, keep

Full Documentation

See comprehensive write-up: interpro/panther/PTHR11857/PTHR11857-summary.md

Immune Gene Pilot Batch Review (6 genes)

Reviewed 6 well-characterized ANOGA immune genes with SPKW-unique annotations to "immune system process" (GO:0002376).

Result: 5/6 (83%) immune annotations are LEGITIMATE

Comparison: D7 vs Immune Genes

Key Pattern: Secondary Over-annotations from Domain Homology

Even when the primary immune annotation is correct, we found secondary over-annotations from automated pipelines:

Biological Insights

1. TEP1: Complement-like opsonin that binds Plasmodium ookinetes via thioester bond
2. SRPN2: Key negative regulator of melanization cascade (inhibits CLIPB9/CLIPB10)
3. PGRPS1: Non-catalytic PGRP that senses bacterial peptidoglycan (Toll pathway)
4. CLIPB4: Upstream protease in prophenoloxidase activation cascade
5. TOLL9: Pattern recognition receptor activating Rel1/NF-κB pathway
6. psidin: NAA25/MDM20 ortholog - NatB complex subunit, NOT an immune protein

PGRP/TEP Extended Batch Review (8 genes)

Extended review of PGRP (peptidoglycan recognition protein) and TEP (thioester-containing protein) families - key innate immune genes in Anopheles.

Result: 8/8 (100%) immune annotations are LEGITIMATE

Key Pattern: PGRP Catalytic vs Non-Catalytic

Key Pattern: TEP "Endopeptidase Inhibitor" Over-annotation

All TEPs (TEP1, TEP2, TEP6, TEP10) are incorrectly annotated with GO:0004866 (endopeptidase inhibitor activity) based on InterPro A2M domain mapping. This is a systematic over-annotation:

• Source: InterPro2GO from IPR001599 (Alpha-2-macroglobulin)
• Problem: Vertebrate A2M proteins ARE protease inhibitors, but insect TEPs have functionally diverged
• Reality: Insect TEPs are complement-like opsonins that covalently tag pathogens via thioester bond
• Action: REMOVE the endopeptidase inhibitor annotation from all ANOGA TEPs

A2M Family Functional Divergence:
┌─────────────────────┐
│   Common Ancestor   │ (A2M domain + thioester)
└──────────┬──────────┘
           │
     ┌─────┴─────┐
     ▼           ▼
┌─────────┐ ┌──────────────┐
│Vertebrate│ │Insect TEPs   │
│A2M       │ │              │
└────┬────┘ └──────┬───────┘
     │             │
  Protease      Complement-like
  Trapping      Opsonization
     │             │
  Inhibitor     Pathogen
  Activity      Tagging

PANTHER Family Analysis: PTHR11857 (OBP/PBP-related)

Family context for D7 proteins: The D7L1 and D7L2 proteins belong to PANTHER family PTHR11857 (Odorant-binding/pheromone-binding protein-related), specifically subfamily SF43 (GEO07291P1-RELATED).

Deep research completed: interpro/panther/PTHR11857/PTHR11857-deep-research-falcon.md

Key findings about the family:

Critical insight - Neo-functionalization

The D7 salivary proteins in subfamily SF43 represent a clear case of neo-functionalization:

• Ancestral function (OBPs): Transport odorants/pheromones to olfactory receptors in sensory lymph
• Derived function (D7s): Sequester host signaling molecules (serotonin, histamine, leukotrienes) in saliva to facilitate blood feeding

Both share:
- Similar structural fold (OBP-like helical architecture)
- Bind similar ligands (biogenic amines)
- Extracellular secretion

But differ in:
- Biological context: Olfactory tissue vs. salivary gland
- Biological role: Sensory perception vs. anti-hemostatic adaptation
- Evolutionary purpose: Chemical communication vs. blood-feeding facilitation

GO annotation implications:

1. "odorant binding" (GO:0005549) - Based on family membership via InterPro, but D7 proteins don't function in olfaction → MODIFY to specific binding terms (serotonin/histamine/leukotriene binding)

2. "toxin activity" (GO:0090729) - Applied via SPKW "Toxin" keyword, but D7 proteins are kratagonists not toxins → OVER-ANNOTATED

3. "extracellular space" (GO:0005615) - Appropriate for both OBPs and D7s (both are secreted)

This case study demonstrates how PANTHER family membership can propagate inappropriate GO annotations when neo-functionalization has occurred.

Recommendations

1. Prioritize immune terms for review: GO:0002376 and GO:0045087 have the most SPKW-unique genes and are highly relevant to vector biology research

2. Validate toxin activity annotations: The D7 proteins suggest some toxin annotations may be legitimate, but "toxin" may not be the best term

3. Check apoptosis/autophagy/rhythm: Based on human experience, expect 80-100% over-annotation rate

4. Consider ANOGA-specific biology: Blood-feeding, olfaction (host-seeking), and innate immunity are core mosquito biology that may have legitimate unique annotations

Review Files Location

genes/ANOGA/
├── [D7r1](../../genes/ANOGA/D7r1/D7r1-ai-review.html)/, [D7r2](../../genes/ANOGA/D7r2/D7r2-ai-review.html)/, [D7r3](../../genes/ANOGA/D7r3/D7r3-ai-review.html)/, [D7r4](../../genes/ANOGA/D7r4/D7r4-ai-review.html)/, [D7r5](../../genes/ANOGA/D7r5/D7r5-ai-review.html)/, [D7L1](../../genes/ANOGA/D7L1/D7L1-ai-review.html)/, [D7L2](../../genes/ANOGA/D7L2/D7L2-ai-review.html)/, D7L3/
├── [TEP1](../../genes/ANOGA/TEP1/TEP1-ai-review.html)/, [TEP2](../../genes/ANOGA/TEP2/TEP2-ai-review.html)/, [TEP6](../../genes/ANOGA/TEP6/TEP6-ai-review.html)/, [TEP10](../../genes/ANOGA/TEP10/TEP10-ai-review.html)/
├── [PGRPS1](../../genes/ANOGA/PGRPS1/PGRPS1-ai-review.html)/, [PGRPLB](../../genes/ANOGA/PGRPLB/PGRPLB-ai-review.html)/, [PGRPLC](../../genes/ANOGA/PGRPLC/PGRPLC-ai-review.html)/, [PGRPLD](../../genes/ANOGA/PGRPLD/PGRPLD-ai-review.html)/, [PGRPS2](../../genes/ANOGA/PGRPS2/PGRPS2-ai-review.html)/, [PGRPS3](../../genes/ANOGA/PGRPS3/PGRPS3-ai-review.html)/
├── [SRPN2](../../genes/ANOGA/SRPN2/SRPN2-ai-review.html)/, [CLIPB4](../../genes/ANOGA/CLIPB4/CLIPB4-ai-review.html)/, [TOLL9](../../genes/ANOGA/TOLL9/TOLL9-ai-review.html)/, [psidin](../../genes/ANOGA/psidin/psidin-ai-review.html)/