User Query
  ↓
Phase 0: CLARIFY + MODE SELECT (factoid vs deep report)
  ↓
Phase 1: TARGET DISAMBIGUATION + PROFILE (default ON for biological targets)
  ├─ Resolve official IDs (Ensembl, UniProt, HGNC)
  ├─ Gather synonyms/aliases + known naming collisions
  ├─ Get protein length, isoforms, domain architecture
  ├─ Get subcellular location, expression, GO terms, pathways
  └─ Output: Target Profile section + Collision-aware search plan
  ↓
Phase 2: LITERATURE SEARCH (internal methodology, not shown)
  ├─ High-precision seed queries (build mechanistic core)
  ├─ Citation network expansion from seeds
  ├─ Collision-filtered broader queries
  └─ Theme clustering + evidence grading
  ↓
Phase 3: REPORT SYNTHESIS
  ├─ Progressive writing to [topic]_report.md
  ├─ Mandatory completeness checklist validation
  └─ Biological model + testable hypotheses
  ↓
Optional: methods_appendix.md (only if user requests)

undefined

**Required verification behavior**:
- Prefer ToolUniverse literature tools (Europe PMC / PubMed / PMC / Semantic Scholar) over general web browsing.
- Use full-text snippet verification when possible (Europe PMC auto-snippet tier is ideal).
- Avoid adding extra claims (e.g., “not X”) unless the paper explicitly supports them.

**Suggested tool pattern**:
- `EuropePMC_search_articles(query=..., extract_terms_from_fulltext=[...])` to pull OA full-text snippets for the key terms.
- If OA snippets unavailable: fall back to `PMC_search_papers` (if in PMC) or `SemanticScholar_search_papers` → `SemanticScholar_get_pdf_snippets`.

**Evidence grading (factoid)**:
- If the statement is explicitly made in a primary experimental paper (Results/Methods/Abstract): label **T1 (★★★)**.
- If it’s only in a review: label **T4 (☆☆☆)** and try to locate the primary source.

UniProt_search → Get UniProt accession for human protein
UniProt_get_entry_by_accession → Full entry with cross-references
UniProt_id_mapping → Map between ID types
ensembl_lookup_gene → Ensembl gene ID, biotype
MyGene_get_gene_annotation → NCBI Gene ID, aliases, summary

undefined

undefined

undefined

undefined

InterPro_get_protein_domains → Domain architecture
UniProt_get_ptm_processing_by_accession → PTMs, active sites
proteins_api_get_protein → Additional protein features

undefined

undefined

HPA_get_subcellular_location → Human Protein Atlas localization
UniProt_get_subcellular_location_by_accession → UniProt annotation

undefined

undefined

GTEx_get_median_gene_expression → Tissue expression (TPM)
HPA_get_rna_expression_by_source → HPA expression data

undefined

undefined

GO_get_annotations_for_gene → GO annotations
Reactome_map_uniprot_to_pathways → Reactome pathways
kegg_get_gene_info → KEGG pathways
OpenTargets_get_target_gene_ontology_by_ensemblID → Open Targets GO

undefined

---

Query 1: "[GENE_SYMBOL]"[Title] AND (mechanism OR function OR structure)
Query 2: "[FULL_PROTEIN_NAME]"[Title] 
Query 3: "[UNIPROT_ID]" (catches supplementary materials)

PubMed_get_cited_by → Papers citing each seed
PubMed_get_related → Computationally related papers  
EuropePMC_get_citations → Alternative citation source
EuropePMC_get_references → Backward citations from seeds

Broader query: "[GENE_SYMBOL]" AND ([pathway1] OR [pathway2] OR [function])
Apply collision filter: NOT [collision_term1] NOT [collision_term2]

"TRAG" AND (T-cell OR immune OR cancer) NOT plasmid NOT conjugation NOT bacterial

results = EuropePMC_search_articles(
    query="bacterial antibiotic resistance evolution",
    limit=10,
    extract_terms_from_fulltext=["ciprofloxacin", "meropenem", "A. baumannii", "MIC"]
)

**Advantages**:
- ✅ Single tool call (search + snippets)
- ✅ Bounded latency (max 3 OA articles, ~3-5 seconds total)
- ✅ No manual URL extraction
- ✅ Max 5 search terms

**Limitations**:
- ❌ Only works for OA articles with fullTextXML
- ❌ Limited to first 3 OA articles
- ❌ Europe PMC coverage only (~30-40% OA)

**When to use**: Initial exploration, quick verification of 1-2 papers

undefined

**ArXiv variant** (100% OA, no paywall):

```python

**Advantages**:
- ✅ Full control over which papers to process
- ✅ Adjustable window size (20-2000 chars)
- ✅ Works for Semantic Scholar (~15-20% OA PDFs) and ArXiv (100%)
- ✅ Can process any number of papers

**Limitations**:
- ❌ Two tool calls per article (search → extract)
- ❌ Manual loop needed
- ❌ Slower than auto-snippet mode

**When to use**: Thorough review of key papers, preprint analysis

undefined

**Limitations**:
- ❌ Requires institutional access
- ❌ No snippet extraction (full HTML)
- ❌ Quality varies by publisher
- ❌ Slowest approach

**When to use**: Last resort for critical paywalled papers

if article.get("open_access") and article.get("fulltext_xml_url"):
    # Proceed with extraction
    pass

if "fulltext_snippets" not in article:
    # Fallback: use abstract or skip
    print(f"No full text available: {article['title']}")

undefined

undefined

Attempt 1: Call tool
If timeout/error:
  Wait 2 seconds
  Attempt 2: Retry
If still fails:
  Wait 5 seconds  
  Attempt 3: Try fallback tool
If fallback fails:
  Document "Data unavailable" in report

*OA Status: Best-effort (Unpaywall not configured)*

undefined

undefined

undefined

---

undefined

---

{
  "metadata": {
    "generated": "2026-02-04",
    "query": "ATP6V1A",
    "total_papers": 342,
    "unique_after_dedup": 287
  },
  "papers": [
    {
      "pmid": "12345678",
      "doi": "10.1038/xxx",
      "title": "Paper Title",
      "authors": ["Smith A", "Jones B"],
      "year": 2024,
      "journal": "Nature",
      "source_databases": ["PubMed", "OpenAlex"],
      "evidence_tier": "T1",
      "themes": ["lysosomal_acidification", "autophagy"],
      "oa_status": "gold",
      "oa_url": "https://...",
      "citation_count": 45,
      "in_core_set": true
    }
  ]
}