Update app.py

app.py

@@ -29,7 +29,7 @@ API_ENDPOINTS = {
     "pubmed": "https://eutils.ncbi.nlm.nih.gov/entrez/eutils/esearch.fcgi",
     "fda_drug_approval": "https://api.fda.gov/drug/label.json",
     "faers_adverse_events": "https://api.fda.gov/drug/event.json",
-    # PharmGKB endpoints (these require a PharmGKB accession such as "PA1234")
+    # PharmGKB endpoints (expecting a PharmGKB accession, e.g. PA1234)
     "pharmgkb_variant_clinical_annotations": "https://api.pharmgkb.org/v1/data/variant/{}/clinicalAnnotations",
     "pharmgkb_gene": "https://api.pharmgkb.org/v1/data/gene/{}",
     "pharmgkb_gene_variants": "https://api.pharmgkb.org/v1/data/gene/{}/variants",
@@ -37,7 +37,7 @@ API_ENDPOINTS = {
     # RxNorm endpoints
     "rxnorm_rxcui": "https://rxnav.nlm.nih.gov/REST/rxcui.json",
     "rxnorm_properties": "https://rxnav.nlm.nih.gov/REST/rxcui/{}/properties.json",
-    # RxClass endpoint (example)
+    # RxClass endpoint
     "rxclass_by_drug": "https://rxnav.nlm.nih.gov/REST/class/byDrugName.json"
 }
 
@@ -123,20 +123,19 @@ def _draw_molecule(smiles: str) -> Optional[Any]:
         return None
 
 def _get_pubchem_drug_details(drug_name: str) -> Optional[Dict[str, str]]:
-    """Retrieves drug details such as molecular formula and IUPAC name from PubChem."""
+    """Retrieves generic drug details (molecular formula, IUPAC name, and canonical SMILES) from PubChem."""
     url = API_ENDPOINTS["pubchem"].format(drug_name)
     data = _query_api(url)
+    details = {}
     if data and data.get("PC_Compounds"):
         compound = data["PC_Compounds"][0]
-        details = {}
         for prop in compound.get("props", []):
             urn = prop.get("urn", {})
             if urn.get("label") == "Molecular Formula":
                 details["Molecular Formula"] = prop["value"]["sval"]
-            elif urn.get("name") in ["Preferred", "Systematic"]:
-                # Use Preferred IUPAC name if available
+            if urn.get("name") == "Preferred":
                 details["IUPAC Name"] = prop["value"]["sval"]
-            elif prop.get("name") == "Canonical SMILES":
+            if prop.get("name") == "Canonical SMILES":
                 details["Canonical SMILES"] = prop["value"]["sval"]
         return details
     return None
@@ -335,30 +334,65 @@ def get_rxclass_by_drug_name(drug_name: str) -> Optional[Dict]:
     return _query_api(url)
 
 # -----------------------------
-# New Function: PubChem Drug Details
+# New Function: Generate AI Insights for a Drug
 # -----------------------------
-def _get_pubchem_drug_details(drug_name: str) -> Optional[Dict[str, str]]:
-    """Retrieves generic drug details (molecular formula, IUPAC name, and canonical SMILES) from PubChem."""
-    url = API_ENDPOINTS["pubchem"].format(drug_name)
-    data = _query_api(url)
-    details = {}
-    if data and data.get("PC_Compounds"):
-        compound = data["PC_Compounds"][0]
-        for prop in compound.get("props", []):
-            urn = prop.get("urn", {})
-            if urn.get("label") == "Molecular Formula":
-                details["Molecular Formula"] = prop["value"]["sval"]
-            if urn.get("name") == "Preferred":
-                details["IUPAC Name"] = prop["value"]["sval"]
-            if prop.get("name") == "Canonical SMILES":
-                details["Canonical SMILES"] = prop["value"]["sval"]
-        return details
-    return None
+def generate_drug_insights(drug_name: str) -> str:
+    """Gathers FDA, PubChem, RxNorm, and RxClass data for a drug and uses GPT‑4 to generate innovative insights."""
+    # FDA Data
+    fda_info = _get_fda_approval(drug_name)
+    fda_status = "Not Approved"
+    if fda_info and fda_info.get("openfda", {}).get("brand_name"):
+        fda_status = ", ".join(fda_info["openfda"]["brand_name"])
+    
+    # PubChem Details
+    pubchem_details = _get_pubchem_drug_details(drug_name)
+    if pubchem_details:
+        formula = pubchem_details.get("Molecular Formula", "N/A")
+        iupac = pubchem_details.get("IUPAC Name", "N/A")
+        canonical_smiles = pubchem_details.get("Canonical SMILES", "N/A")
+    else:
+        formula = iupac = canonical_smiles = "Not Available"
+    
+    # RxNorm Data
+    rxnorm_id = get_rxnorm_rxcui(drug_name)
+    if rxnorm_id:
+        rx_properties = get_rxnorm_properties(rxnorm_id)
+        rxnorm_info = f"RxCUI: {rxnorm_id}. Properties: {rx_properties}"
+    else:
+        rxnorm_info = "No RxNorm data available."
+    
+    # RxClass Data
+    rxclass_data = get_rxclass_by_drug_name(drug_name)
+    if rxclass_data and rxclass_data.get("classMember"):
+        rxclass_info = f"RxClass: {rxclass_data}"
+    else:
+        rxclass_info = "No RxClass data available."
+    
+    # Construct prompt for GPT‑4 with all the gathered data
+    prompt = (
+        f"Drug Analysis Report for '{drug_name}':\n\n"
+        f"**FDA Approval Status:** {fda_status}\n\n"
+        f"**PubChem Details:**\n"
+        f"  - Molecular Formula: {formula}\n"
+        f"  - IUPAC Name: {iupac}\n"
+        f"  - Canonical SMILES: {canonical_smiles}\n\n"
+        f"**RxNorm Data:** {rxnorm_info}\n\n"
+        f"**RxClass Data:** {rxclass_info}\n\n"
+        f"As an advanced pharmacogenomics researcher and AI expert, please provide an innovative and comprehensive analysis of "
+        f"the drug '{drug_name}'. In your response, include:\n"
+        f"- Pharmacogenomic considerations\n"
+        f"- Potential repurposing opportunities\n"
+        f"- Regulatory insights and challenges\n"
+        f"- Suggestions for further research and data integration\n\n"
+        f"Present your answer in a clear, bullet-point format and feel free to add any novel ideas."
+    )
+    
+    insights = generate_content(prompt)
+    return insights
 
 # -----------------------------
 # Streamlit App Layout and Tabs
 # -----------------------------
-# Define tabs including a new "Drug Data Integration" tab.
 tabs = st.tabs([
     "💊 Drug Development",
     "📊 Trial Analytics",
@@ -366,7 +400,8 @@ tabs = st.tabs([
     "📜 Regulatory Intelligence",
     "📚 Literature Search",
     "📈 Dashboard",
-    "🧪 Drug Data Integration"
+    "🧪 Drug Data Integration",
+    "🤖 AI Insights"
 ])
 
 # -----------------------------
@@ -505,7 +540,7 @@ with tabs[2]:
 # -----------------------------
 with tabs[3]:
     st.header("Global Regulatory Monitoring")
-    st.markdown("**Note:** EMA, WHO, and DailyMed endpoints have been removed due to persistent errors. Instead, we provide FDA data and generic drug details (formula, IUPAC name, structure) from PubChem.")
+    st.markdown("**Note:** Due to persistent issues with EMA, WHO, and DailyMed APIs, this section now focuses on FDA data and generic drug details from PubChem.")
     drug_name = st.text_input("Drug Product:", placeholder="Enter generic or brand name")
     
     if st.button("Generate Regulatory Report"):
@@ -521,27 +556,27 @@ with tabs[3]:
             if pubchem_details:
                 formula = pubchem_details.get("Molecular Formula", "N/A")
                 iupac = pubchem_details.get("IUPAC Name", "N/A")
-                smiles = pubchem_details.get("Canonical SMILES", "N/A")
+                canonical_smiles = pubchem_details.get("Canonical SMILES", "N/A")
             else:
-                formula = iupac = smiles = "Not Available"
+                formula = iupac = canonical_smiles = "Not Available"
             
-            st.subheader("Regulatory Status")
+            st.subheader("Regulatory Status & Drug Details")
             col1, col2 = st.columns(2)
             with col1:
                 st.markdown("**FDA Status**")
                 st.write(fda_status)
             with col2:
-                st.markdown("**Generic/Formula Details (PubChem)**")
+                st.markdown("**Drug Details (PubChem)**")
                 st.write(f"**Molecular Formula:** {formula}")
                 st.write(f"**IUPAC Name:** {iupac}")
-                st.write(f"**Canonical SMILES:** {smiles}")
+                st.write(f"**Canonical SMILES:** {canonical_smiles}")
             
             regulatory_content = (
-                f"### Regulatory Report\n\n"
+                f"### Regulatory Report for {drug_name}\n\n"
                 f"**FDA Status:** {fda_status}\n\n"
                 f"**Molecular Formula:** {formula}\n\n"
                 f"**IUPAC Name:** {iupac}\n\n"
-                f"**Canonical SMILES:** {smiles}\n"
+                f"**Canonical SMILES:** {canonical_smiles}\n"
             )
             with tempfile.NamedTemporaryFile(delete=False, suffix=".pdf") as tmp_file:
                 report_file = _save_pdf_report(regulatory_content, tmp_file.name)
@@ -595,8 +630,6 @@ with tabs[5]:
     
     # Placeholder KPI counts (replace with real aggregated data if available)
     fda_count = 5000       # Example value
-    ema_count = 3000       # Example value (not used now)
-    who_count = 1500       # Example value (not used now)
     trials_count = 12000   # Example value
     pub_count = 250000     # Example value
     
@@ -666,7 +699,7 @@ with tabs[6]:
             else:
                 st.write("No RxClass data found for the given drug.")
             
-            # Additionally, show generic drug details from PubChem
+            # PubChem Drug Details for generic information
             pubchem_details = _get_pubchem_drug_details(drug_query)
             st.subheader("PubChem Drug Details")
             if pubchem_details:
@@ -676,4 +709,16 @@ with tabs[6]:
             else:
                 st.write("No PubChem details found for the given drug.")
 
+# -----------------------------
+# Tab 8: AI Insights
+# -----------------------------
+with tabs[7]:
+    st.header("🤖 AI Insights")
+    ai_drug_query = st.text_input("Enter Drug Name for AI-Driven Analysis:", placeholder="e.g., aspirin")
+    if st.button("Generate AI Insights"):
+        with st.spinner("Generating AI insights..."):
+            insights = generate_drug_insights(ai_drug_query)
+            st.subheader("AI-Driven Drug Analysis")
+            st.markdown(insights)
+