""" BioPrime Molecular Docking Demo ================================ Interactive demo showcasing AI-powered molecular docking for drug discovery. Features: - 3D protein structure visualization - Compound library selection - Docking simulation with binding energy results - BSV blockchain verification display Powered by: Origin Neural AI Docking Engine Sponsored by: Smartledger Solutions, Origin Neural AI, Bryan Daugherty Website: https://bioprime.one """ import gradio as gr import requests import json import hashlib import random from datetime import datetime from typing import Optional, Dict, List, Tuple import time # ============================================================================= # Configuration # ============================================================================= BIOPRIME_API = "https://bioprime.one/api/v1" RCSB_PDB_URL = "https://files.rcsb.org/download" # Demo targets with sponsored research campaigns DEMO_TARGETS = [ { "id": "melanoma", "name": "BRAF V600E - Melanoma", "pdb": "4MNE", "sponsor": "Bryan Daugherty", "disease": "Melanoma", "description": "Mutated BRAF kinase found in ~50% of melanomas, target for vemurafenib-like inhibitors.", "binding_site": {"x": 25.0, "y": 5.0, "z": 15.0}, "color": "#FF6B6B" }, { "id": "diabetes", "name": "DPP-4 - Type 2 Diabetes", "pdb": "2ONC", "sponsor": "Bryan Daugherty", "disease": "Type 2 Diabetes", "description": "Dipeptidyl peptidase-4, target for incretin-based diabetes medications like sitagliptin.", "binding_site": {"x": 35.0, "y": 40.0, "z": 45.0}, "color": "#4ECDC4" }, { "id": "covid", "name": "COVID-19 Main Protease", "pdb": "6LU7", "sponsor": "BioPrime Community", "disease": "COVID-19", "description": "SARS-CoV-2 main protease (Mpro), essential for viral replication. Target for Paxlovid.", "binding_site": {"x": -10.8, "y": 35.2, "z": 63.4}, "color": "#9B59B6" }, { "id": "hiv", "name": "HIV-1 Protease", "pdb": "1HVR", "sponsor": "Origin Neural AI", "disease": "HIV/AIDS", "description": "Critical enzyme for HIV replication, target for protease inhibitor antiretroviral drugs.", "binding_site": {"x": -6.2, "y": 20.1, "z": 41.8}, "color": "#E74C3C" }, { "id": "lung", "name": "EGFR Kinase - Lung Cancer", "pdb": "1M17", "sponsor": "Smartledger & Origin Neural AI", "disease": "Non-small Cell Lung Cancer", "description": "Epidermal growth factor receptor, key target in NSCLC therapy. Target for erlotinib.", "binding_site": {"x": 40.5, "y": 0.6, "z": 56.0}, "color": "#3498DB" }, { "id": "breast", "name": "CDK4/6 - Breast Cancer", "pdb": "5L2I", "sponsor": "Smartledger", "disease": "Breast Cancer", "description": "Cyclin-dependent kinase 4/6 inhibitor target for hormone-receptor positive breast cancer.", "binding_site": {"x": 15.0, "y": 25.0, "z": 35.0}, "color": "#E91E63" }, ] # Demo compound library COMPOUND_LIBRARY = [ {"id": "aspirin", "name": "Aspirin", "smiles": "CC(=O)OC1=CC=CC=C1C(=O)O", "mw": 180.16, "category": "Anti-inflammatory"}, {"id": "ibuprofen", "name": "Ibuprofen", "smiles": "CC(C)CC1=CC=C(C=C1)C(C)C(=O)O", "mw": 206.29, "category": "Anti-inflammatory"}, {"id": "caffeine", "name": "Caffeine", "smiles": "CN1C=NC2=C1C(=O)N(C(=O)N2C)C", "mw": 194.19, "category": "Stimulant"}, {"id": "paracetamol", "name": "Acetaminophen", "smiles": "CC(=O)NC1=CC=C(O)C=C1", "mw": 151.16, "category": "Analgesic"}, {"id": "metformin", "name": "Metformin", "smiles": "CN(C)C(=N)NC(=N)N", "mw": 129.17, "category": "Antidiabetic"}, {"id": "atorvastatin", "name": "Atorvastatin", "smiles": "CC(C)C1=C(C(=C(N1CCC(CC(CC(=O)O)O)O)C2=CC=C(C=C2)F)C3=CC=CC=C3)C(=O)NC4=CC=CC=C4", "mw": 558.64, "category": "Statin"}, {"id": "nirmatrelvir", "name": "Nirmatrelvir", "smiles": "CC1(CC1)C(=O)NC(CC2CCNC2=O)C(=O)NC(CC(F)(F)F)C#N", "mw": 499.53, "category": "Antiviral (COVID-19)"}, {"id": "oseltamivir", "name": "Oseltamivir", "smiles": "CCOC(=O)C1=CC(OC(CC)CC)C(NC(C)=O)C(N)C1", "mw": 312.41, "category": "Antiviral (Flu)"}, {"id": "remdesivir", "name": "Remdesivir", "smiles": "CCC(CC)COC(=O)C(C)NP(=O)(OCC1C(C(C(O1)N2C=CC(=O)NC2=O)O)O)OC3=CC=CC=C3", "mw": 602.58, "category": "Antiviral"}, {"id": "sitagliptin", "name": "Sitagliptin", "smiles": "NC(CC(=O)N1CCN2C(C1)=NN=C2C(F)(F)F)CC1=C(F)C=C(F)C(F)=C1F", "mw": 407.31, "category": "Antidiabetic (DPP-4)"}, {"id": "vemurafenib", "name": "Vemurafenib", "smiles": "CCCS(=O)(=O)NC1=CC=C(C=C1)C2=NC(=C(S2)C3=CC(=NC=C3)NC4=CC=C(C=C4)Cl)C#N", "mw": 489.93, "category": "Kinase Inhibitor (Melanoma)"}, {"id": "erlotinib", "name": "Erlotinib", "smiles": "COCCOC1=C(C=C2C(=C1)C(=NC=N2)NC3=CC(=C(C=C3)F)Cl)OCCOC", "mw": 393.44, "category": "EGFR Inhibitor"}, ] # Pre-computed docking results (simulated but realistic) PRECOMPUTED_RESULTS = { "melanoma": { "vemurafenib": -9.8, "erlotinib": -7.2, "caffeine": -4.1, "aspirin": -5.3, "ibuprofen": -5.8, }, "diabetes": { "sitagliptin": -10.2, "metformin": -6.8, "caffeine": -4.5, "aspirin": -4.9, "ibuprofen": -5.1, }, "covid": { "nirmatrelvir": -8.9, "remdesivir": -7.6, "caffeine": -5.2, "aspirin": -4.8, "oseltamivir": -6.4, }, "hiv": { "remdesivir": -7.8, "oseltamivir": -6.2, "caffeine": -4.3, "aspirin": -4.5, "atorvastatin": -6.9, }, "lung": { "erlotinib": -9.4, "vemurafenib": -7.1, "caffeine": -4.0, "aspirin": -4.7, "atorvastatin": -6.5, }, "breast": { "atorvastatin": -7.3, "erlotinib": -6.8, "caffeine": -4.2, "aspirin": -4.4, "metformin": -5.1, }, } # ============================================================================= # Helper Functions # ============================================================================= def fetch_pdb_structure(pdb_id: str) -> Optional[str]: """Fetch PDB structure from RCSB.""" try: # Try BioPrime API first response = requests.get(f"{BIOPRIME_API}/docking/demo/pdb/{pdb_id}", timeout=10) if response.status_code == 200: data = response.json() return data.get("pdb_content", data.get("pdb_data", None)) except: pass # Fallback to RCSB try: response = requests.get(f"{RCSB_PDB_URL}/{pdb_id}.pdb", timeout=10) if response.status_code == 200: return response.text except: pass return None def create_3d_viewer(pdb_content: str, binding_site: dict = None, style: str = "cartoon") -> str: """Create 3D molecular viewer HTML using iframe with srcdoc for JS execution.""" import html import base64 # Escape PDB content for JavaScript (double escape for iframe) pdb_escaped = pdb_content.replace('\\', '\\\\').replace('\n', '\\n').replace('\r', '').replace("'", "\\'").replace('"', '\\"') # Build style configuration if style == "cartoon": style_js = "viewer.setStyle({}, {cartoon: {color: 'spectrum'}});" elif style == "surface": style_js = "viewer.setStyle({}, {cartoon: {color: 'spectrum'}}); viewer.addSurface($3Dmol.SAS, {opacity: 0.7, color: 'white'});" elif style == "stick": style_js = "viewer.setStyle({}, {stick: {colorscheme: 'Jmol'}});" elif style == "sphere": style_js = "viewer.setStyle({}, {sphere: {colorscheme: 'Jmol', scale: 0.3}});" else: style_js = "viewer.setStyle({}, {cartoon: {color: 'spectrum'}});" # Binding site sphere binding_site_js = "" if binding_site: binding_site_js = f"viewer.addSphere({{center: {{x: {binding_site['x']}, y: {binding_site['y']}, z: {binding_site['z']}}}, radius: 8, color: 'red', opacity: 0.3}});" # Create complete HTML document for iframe iframe_html = f'''
''' # Escape for srcdoc attribute iframe_srcdoc = html.escape(iframe_html) html_content = f'''
''' return html_content def generate_docking_result(target_id: str, compound_ids: List[str]) -> Tuple[str, str, str]: """ Simulate docking and return results. Returns: (results_text, binding_chart_data, receipt) """ target = next((t for t in DEMO_TARGETS if t["id"] == target_id), None) if not target: return "Target not found", "", "" results = [] precomputed = PRECOMPUTED_RESULTS.get(target_id, {}) for comp_id in compound_ids: compound = next((c for c in COMPOUND_LIBRARY if c["id"] == comp_id), None) if not compound: continue # Use precomputed or generate realistic random if comp_id in precomputed: energy = precomputed[comp_id] else: # Generate realistic binding energy based on molecular weight base_energy = -4.0 - (compound["mw"] / 100) energy = round(base_energy + random.uniform(-1.5, 1.5), 2) results.append({ "compound": compound["name"], "smiles": compound["smiles"], "energy": energy, "mw": compound["mw"], "category": compound["category"] }) # Sort by binding energy (more negative = better) results.sort(key=lambda x: x["energy"]) # Generate results text results_text = f""" ## Docking Results for {target['name']} **Target Disease:** {target['disease']} **Sponsor:** {target['sponsor']} **PDB ID:** {target['pdb']} --- ### Top Binding Compounds | Rank | Compound | Binding Energy | Category | |------|----------|----------------|----------| """ for i, r in enumerate(results[:10], 1): emoji = "🏆" if i == 1 else "đŸĨˆ" if i == 2 else "đŸĨ‰" if i == 3 else " " results_text += f"| {emoji} {i} | **{r['compound']}** | {r['energy']:.2f} kcal/mol | {r['category']} |\n" results_text += f""" --- ### Interpretation - **Best Hit:** {results[0]['compound']} with {results[0]['energy']:.2f} kcal/mol - **Binding energies < -7 kcal/mol** indicate strong binding potential - **Binding energies < -9 kcal/mol** suggest drug-like affinity --- *Powered by Origin Neural AI Docking Engine* *Results simulated for demonstration - actual BioPrime uses GPU-accelerated physics* """ # Generate chart data chart_labels = [r["compound"][:12] for r in results[:8]] chart_values = [abs(r["energy"]) for r in results[:8]] chart_html = f"""

Binding Affinity Comparison

""" max_val = max(chart_values) if chart_values else 1 colors = ["#FF6B6B", "#4ECDC4", "#45B7D1", "#96CEB4", "#FFEAA7", "#DDA0DD", "#98D8C8", "#F7DC6F"] for i, (label, val) in enumerate(zip(chart_labels, chart_values)): height = int((val / max_val) * 150) color = colors[i % len(colors)] chart_html += f"""
-{val:.1f}
{label}
""" chart_html += """

Binding Energy (kcal/mol) - Higher bars = stronger binding

""" # Add social sharing buttons share_text = f"I just screened compounds against {target['name']} using BioPrime! Best hit: {results[0]['compound']} at {results[0]['energy']:.2f} kcal/mol. Try AI-powered drug discovery:" share_url = "https://bioprime.one" import urllib.parse encoded_text = urllib.parse.quote(share_text) encoded_url = urllib.parse.quote(share_url) chart_html += f"""

Share Your Discovery

Post on X LinkedIn Facebook

Want real blockchain-verified results? Sign up at bioprime.one

""" # Generate blockchain receipt job_id = f"DEMO-{target_id.upper()}-{hashlib.md5(str(compound_ids).encode()).hexdigest()[:8]}" data_hash = hashlib.sha256(json.dumps(results, sort_keys=True).encode()).hexdigest() receipt = generate_demo_receipt( job_id=job_id, target_name=target['name'], compounds_screened=len(compound_ids), top_hits=len(results), best_energy=results[0]['energy'] if results else 0, data_hash=data_hash ) return results_text, chart_html, receipt def generate_demo_receipt(job_id: str, target_name: str, compounds_screened: int, top_hits: int, best_energy: float, data_hash: str) -> str: """Generate a demo blockchain receipt.""" timestamp = datetime.utcnow().strftime("%Y-%m-%d %H:%M:%S UTC") demo_txid = f"demo_{hashlib.md5(data_hash.encode()).hexdigest()[:48]}" receipt = f""" ╔══════════════════════════════════════════════════════════════════════════╗ ║ BIOPRIME DISCOVERY CERTIFICATE ║ ║ [DEMO - NOT ON CHAIN] ║ ╠══════════════════════════════════════════════════════════════════════════â•Ŗ ║ ║ ║ JOB ID: {job_id:<57} ║ ║ TARGET: {target_name[:47]:<57} ║ ║ ║ ╠══════════════════════════════════════════════════════════════════════════â•Ŗ ║ SCREENING RESULTS ║ ╠══════════════════════════════════════════════════════════════════════════â•Ŗ ║ ║ ║ Compounds Screened: {compounds_screened:<45} ║ ║ Top Poses Generated: {top_hits:<45} ║ ║ Best Binding Energy: {best_energy:.2f} kcal/mol{' ':<36} ║ ║ ║ ╠══════════════════════════════════════════════════════════════════════════â•Ŗ ║ VERIFICATION DETAILS ║ ╠══════════════════════════════════════════════════════════════════════════â•Ŗ ║ ║ ║ Timestamp: {timestamp:<53} ║ ║ Data Hash: sha256:{data_hash[:49]:<46} ║ ║ ║ ╠══════════════════════════════════════════════════════════════════════════â•Ŗ ║ BLOCKCHAIN VERIFICATION ║ ╠══════════════════════════════════════════════════════════════════════════â•Ŗ ║ ║ ║ Network: BSV (Bitcoin SV) - Demo Mode ║ ║ Status: Demo certificate - not anchored to blockchain ║ ║ ║ ║ For real blockchain-verified results, visit: ║ ║ https://bioprime.one ║ ║ ║ ╠══════════════════════════════════════════════════════════════════════════â•Ŗ ║ ║ ║ BioPrime anchors real docking results to the BSV blockchain for ║ ║ immutable proof of discovery. Sign up to run verified experiments. ║ ║ ║ ║ ━━━ bioprime.one ━━━ ║ ║ ║ ╚══════════════════════════════════════════════════════════════════════════╝ """ return receipt.strip() # ============================================================================= # Gradio Interface # ============================================================================= def view_protein(target_name: str, view_style: str) -> Tuple[str, str]: """View selected protein structure.""" target = next((t for t in DEMO_TARGETS if t["name"] == target_name), None) if not target: return "

Please select a target

", "" pdb_content = fetch_pdb_structure(target["pdb"]) if not pdb_content: return f"

Failed to fetch PDB structure for {target['pdb']}

", "" viewer_html = create_3d_viewer(pdb_content, target.get("binding_site"), view_style.lower()) info_html = f"""

{target['name']}

PDB ID: {target['pdb']}

Disease: {target['disease']}

Sponsor: {target['sponsor']}

{target['description']}

💡 The red sphere indicates the active binding site where drug candidates interact with the protein.

""" return viewer_html, info_html def run_docking(target_name: str, compounds: List[str]) -> Tuple[str, str, str]: """Run docking simulation.""" if not target_name: return "Please select a target protein", "", "" if not compounds: return "Please select at least one compound", "", "" target = next((t for t in DEMO_TARGETS if t["name"] == target_name), None) if not target: return "Invalid target", "", "" # Get compound IDs from names compound_ids = [] for comp_name in compounds: comp = next((c for c in COMPOUND_LIBRARY if c["name"] == comp_name), None) if comp: compound_ids.append(comp["id"]) # Simulate docking (small delay for effect) time.sleep(1.5) results_text, chart_html, receipt = generate_docking_result(target["id"], compound_ids) return results_text, chart_html, receipt # Create the Gradio interface with gr.Blocks( title="BioPrime Molecular Docking Demo", theme=gr.themes.Base( primary_hue="teal", secondary_hue="purple", neutral_hue="slate", font=gr.themes.GoogleFont("Inter") ), css=""" .gradio-container { max-width: 1400px !important; background: linear-gradient(135deg, #0f0c29 0%, #302b63 50%, #24243e 100%) !important; } .gr-button-primary { background: linear-gradient(135deg, #4ECDC4 0%, #44A08D 100%) !important; } .header-text { text-align: center; color: white; } footer {display: none !important;} """ ) as demo: # Header gr.HTML("""

đŸ§Ŧ BioPrime

AI-Powered Molecular Docking for Drug Discovery

10,000x Faster â€ĸ $5 per Million Compounds â€ĸ Blockchain-Verified Results

""") with gr.Tabs(): # Tab 1: Interactive Docking Demo with gr.TabItem("đŸ”Ŧ Try Docking", id="docking"): gr.Markdown(""" ### Dock Drug Candidates Against Disease Targets Select a protein target and compounds to simulate molecular docking. See binding energies and get a blockchain-ready certificate. """) with gr.Row(): with gr.Column(scale=1): target_dropdown = gr.Dropdown( choices=[t["name"] for t in DEMO_TARGETS], label="đŸŽ¯ Select Disease Target", info="Choose from sponsored research targets" ) compound_select = gr.CheckboxGroup( choices=[c["name"] for c in COMPOUND_LIBRARY], label="💊 Select Compounds to Test", info="Choose multiple compounds for screening" ) dock_btn = gr.Button("🚀 Run Docking Simulation", variant="primary", size="lg") gr.HTML("""

💡 Quick Start

  1. Select BRAF V600E - Melanoma
  2. Check Vemurafenib (the actual drug!)
  3. Add a few other compounds to compare
  4. Click Run Docking
""") with gr.Column(scale=2): results_md = gr.Markdown("*Results will appear here after docking...*") chart_html = gr.HTML() with gr.Accordion("📜 Blockchain Certificate (Demo)", open=False): receipt_text = gr.Code(label="Discovery Certificate", language=None, lines=30) dock_btn.click( fn=run_docking, inputs=[target_dropdown, compound_select], outputs=[results_md, chart_html, receipt_text] ) # Tab 2: 3D Protein Viewer with gr.TabItem("🔮 3D Protein Viewer", id="viewer"): gr.Markdown(""" ### Explore Protein Structures in 3D Visualize the molecular targets for drug discovery. The red sphere indicates the binding site. """) with gr.Row(): with gr.Column(scale=1): viewer_target = gr.Dropdown( choices=[t["name"] for t in DEMO_TARGETS], label="đŸŽ¯ Select Protein", value=DEMO_TARGETS[0]["name"] ) view_style = gr.Radio( choices=["Cartoon", "Surface", "Stick", "Sphere"], value="Cartoon", label="🎨 Visualization Style" ) view_btn = gr.Button("👁ī¸ View Structure", variant="primary") target_info = gr.HTML() with gr.Column(scale=2): viewer_output = gr.HTML( value="

Select a protein and click 'View Structure'

" ) view_btn.click( fn=view_protein, inputs=[viewer_target, view_style], outputs=[viewer_output, target_info] ) # Tab 3: About BioPrime with gr.TabItem("ℹī¸ About", id="about"): gr.Markdown(""" ## About BioPrime Network BioPrime is a **decentralized molecular docking platform** that makes drug discovery accessible to everyone. ### Key Features | Feature | Traditional | BioPrime | |---------|-------------|----------| | Speed | Days-Weeks | Minutes | | Cost | $10,000+ | $5/million | | Verification | Manual | Blockchain | | Access | Limited | Open | ### How It Works 1. **Submit** - Upload your protein target or select from our library 2. **Screen** - Our Origin Neural AI engine docks millions of compounds 3. **Discover** - Get ranked binding poses with energy scores 4. **Verify** - Results anchored to BSV blockchain for immutable proof ### Sponsored Research Campaigns BioPrime features sponsored research targets where community members can contribute to drug discovery: - **Bryan Daugherty** - Melanoma (BRAF V600E), Type 2 Diabetes (DPP-4) - **Smartledger** - Breast Cancer (CDK4/6) - **Origin Neural AI** - HIV-1 Protease - **Greg Ward** - Tuberculosis, Dengue Fever - **Shawn Ryan** - Alzheimer's, Parkinson's ### Technology Stack - **Docking Engine**: Origin Neural AI (GPU-accelerated) - **Blockchain**: BSV (Bitcoin SV) for immutable verification - **Backend**: FastAPI, Python - **Frontend**: React, TypeScript --- ### Get Started **🌐 Visit [bioprime.one](https://bioprime.one) to run real docking experiments!** - Sign up for free (10 free credits) - Screen up to 1 million compounds per job - Get blockchain-verified discovery certificates - Participate in sponsored research campaigns ---

Ready to discover the next breakthrough drug?

🚀 Launch BioPrime →

""") # Footer gr.HTML("""

bioprime.one

Powered by Origin Neural AI â€ĸ Blockchain verification on BSV

""") if __name__ == "__main__": demo.launch()