feat: add Europe PMC source + update STATUS.md

STATUS.md

@@ -6,7 +6,7 @@ _Last updated: May 4, 2026_
 Multi-agent pipeline in agent/agent.py with 5 stages:
 
 1. Query Architect: generates 5 MeSH-optimised PubMed queries via Qwen2.5-72B on AMD MI300X
-2. Literature Scout: fetches all queries sequentially with rate limiting
+2. Literature Scout: fetches from PubMed and Europe PMC in parallel, deduplicates by PMID
 3. PRISMA Filter: automatic inclusion/exclusion screening with one-line reasons, user can override any decision
 4. Evidence Synthesiser: structured synthesis with Background, Key Findings, Level of Evidence, Conflicting Evidence, Research Gaps, Clinical Implications
 5. Citation Builder: formatted references with PMID, synthesis runs on PRISMA-included papers only
@@ -25,7 +25,7 @@ Additional features completed:
 * Query refinement suggestions: 3 AI-generated follow-up research questions based on synthesis gaps
 * Session history: queries saved to sessions.json, reloadable from sidebar
 * Rate limit retry logic: automatic backoff on API errors
-* SSL patch for PubMed Entrez on corporate/university networks
+* SSL patch for PubMed and Europe PMC Entrez on corporate/university networks
 * Signature: Azlaan Mohammad 2026 in footer
 
 ## UI Updates (May 4 2026)
@@ -48,7 +48,7 @@ Additional features completed:
 
 * LLM: Qwen2.5-72B-Instruct on AMD MI300X via vLLM 0.17.1
 * Agent Framework: LangGraph + LangChain
-* Literature Retrieval: BioPython Entrez / PubMed NCBI
+* Literature Retrieval: BioPython Entrez / PubMed NCBI + Europe PMC
 * Web Framework: Flask with SSE streaming
 * PDF: ReportLab
 * Frontend: HTML, CSS, vanilla JS

agent/agent.py

@@ -64,13 +64,16 @@ def run_query_architect(user_question):
 
 
 def run_literature_scout(queries):
+    from retrieval.pubmed import fetch_europepmc
     all_papers = {}
-    def fetch_one(q):
-        return fetch_pubmed(q, max_results=5)
     import time
     for q in queries:
         time.sleep(0.4)
-        for r in fetch_one(q):
+        for r in fetch_pubmed(q, max_results=5):
+            if r["pmid"] not in all_papers:
+                all_papers[r["pmid"]] = r
+        time.sleep(0.4)
+        for r in fetch_europepmc(q, max_results=3):
             if r["pmid"] not in all_papers:
                 all_papers[r["pmid"]] = r
     return all_papers

retrieval/pubmed.py

@@ -87,3 +87,45 @@ if __name__ == "__main__":
         print(f"Journal: {r['journal']}")
         print(f"Year:    {r['year']}")
         print("-" * 60)
+
+def fetch_europepmc(query: str, max_results: int = 5) -> list:
+    import ssl, urllib.request, urllib.parse, json
+    ctx = ssl.create_default_context()
+    ctx.check_hostname = False
+    ctx.verify_mode = ssl.CERT_NONE
+    params = urllib.parse.urlencode({
+        "query": query,
+        "resultType": "core",
+        "pageSize": max_results,
+        "format": "json",
+        "sort": "CITED desc"
+    })
+    url = "https://www.ebi.ac.uk/europepmc/webservices/rest/search?" + params
+    try:
+        with urllib.request.urlopen(url, context=ctx, timeout=20) as r:
+            data = json.loads(r.read())
+        results = []
+        for p in data.get("resultList", {}).get("result", []):
+            pmid = p.get("pmid", p.get("id", ""))
+            if not pmid:
+                continue
+            authors_list = []
+            for a in p.get("authorList", {}).get("author", []):
+                name = a.get("fullName", "")
+                if name:
+                    authors_list.append(name)
+            authors = ", ".join(authors_list[:3])
+            if len(authors_list) > 3:
+                authors += " et al."
+            results.append({
+                "pmid": str(pmid),
+                "title": p.get("title", "Title unavailable").rstrip("."),
+                "authors": authors or "Authors unavailable",
+                "journal": p.get("journalTitle", "Journal unavailable"),
+                "year": str(p.get("pubYear", "n.d.")),
+                "abstract": p.get("abstractText", "Abstract not available")
+            })
+        return results
+    except Exception as e:
+        print(f"[EuropePMC] fetch failed: {e}")
+        return []