Spaces:
Running
Running
File size: 17,396 Bytes
0efdc2f |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 |
#!/usr/bin/env python3
"""
Demo: Full Stack DeepCritical Agent (Phases 1-8).
This script demonstrates the COMPLETE drug repurposing research pipeline:
- Phase 2: Search (PubMed + Web)
- Phase 6: Embeddings (Semantic deduplication)
- Phase 7: Hypothesis (Mechanistic reasoning)
- Phase 3: Judge (Evidence assessment)
- Phase 8: Report (Structured scientific report)
Usage:
# Full demo with real searches and LLM (requires API keys)
uv run python examples/full_stack_demo/run_full.py "metformin Alzheimer's"
# Mock mode - demonstrates pipeline without API calls
uv run python examples/full_stack_demo/run_full.py --mock
# With specific iterations
uv run python examples/full_stack_demo/run_full.py "sildenafil heart failure" --iterations 2
"""
import argparse
import asyncio
import os
import sys
from typing import Any
from src.utils.models import Citation, Evidence, MechanismHypothesis
def print_header(title: str) -> None:
"""Print a formatted section header."""
print(f"\n{'='*70}")
print(f" {title}")
print(f"{'='*70}\n")
def print_step(step: int, name: str) -> None:
"""Print a step indicator."""
print(f"\n[Step {step}] {name}")
print("-" * 50)
def create_mock_evidence() -> list[Evidence]:
"""Create comprehensive mock evidence for demo without API calls."""
return [
Evidence(
content=(
"Metformin, a first-line treatment for type 2 diabetes, activates "
"AMP-activated protein kinase (AMPK). AMPK is a master metabolic "
"regulator that inhibits mTOR signaling, reducing protein synthesis "
"and cell proliferation. This mechanism has implications beyond "
"glucose control."
),
citation=Citation(
source="pubmed",
title="Metformin activates AMPK through LKB1-dependent mechanisms",
url="https://pubmed.ncbi.nlm.nih.gov/19001324/",
date="2023-06",
authors=["Zhang L", "Wang H", "Chen Y"],
),
),
Evidence(
content=(
"In transgenic mouse models of Alzheimer's disease, metformin treatment "
"reduced tau phosphorylation by 45% and decreased amyloid-beta plaque "
"formation. Treated mice showed improved performance on Morris water "
"maze tests, suggesting preserved spatial memory."
),
citation=Citation(
source="pubmed",
title="Metformin ameliorates tau pathology in AD mouse models",
url="https://pubmed.ncbi.nlm.nih.gov/31256789/",
date="2024-01",
authors=["Kim J", "Lee S", "Park M", "Tanaka K"],
),
),
Evidence(
content=(
"A population-based cohort study of 100,000 diabetic patients found "
"that metformin users had 35% lower risk of developing Alzheimer's "
"disease compared to sulfonylurea users (HR=0.65, 95% CI: 0.58-0.73). "
"The protective effect increased with duration of use."
),
citation=Citation(
source="pubmed",
title="Metformin and dementia risk: UK Biobank analysis",
url="https://pubmed.ncbi.nlm.nih.gov/34567890/",
date="2023-09",
authors=["Smith A", "Johnson B", "Williams C"],
),
),
Evidence(
content=(
"mTOR hyperactivation is observed in Alzheimer's disease brain tissue. "
"mTOR inhibition by rapamycin or metformin promotes autophagy, which "
"clears misfolded proteins including tau and amyloid-beta aggregates. "
"This suggests a common therapeutic pathway."
),
citation=Citation(
source="pubmed",
title="mTOR-autophagy axis in neurodegeneration",
url="https://pubmed.ncbi.nlm.nih.gov/32109876/",
date="2023-03",
authors=["Brown C", "Davis D", "Miller E"],
),
),
Evidence(
content=(
"Metformin crosses the blood-brain barrier via organic cation "
"transporters (OCT1, OCT2). CSF concentrations reach approximately "
"1-2% of plasma levels, sufficient for AMPK activation in neurons. "
"Brain accumulation is observed in hippocampus and prefrontal cortex."
),
citation=Citation(
source="pubmed",
title="Brain pharmacokinetics of metformin in humans",
url="https://pubmed.ncbi.nlm.nih.gov/35678901/",
date="2024-02",
authors=["Wilson E", "Garcia F"],
),
),
Evidence(
content=(
"Phase 2 clinical trial (NCT04098666) showed metformin 2000mg/day "
"for 12 months slowed cognitive decline by 18% compared to placebo "
"in patients with mild cognitive impairment. Biomarker analysis "
"showed reduced CSF tau levels in the treatment group."
),
citation=Citation(
source="web",
title="Metformin for Alzheimer's prevention trial results",
url="https://clinicaltrials.gov/ct2/show/NCT04098666",
date="2024-03",
authors=["NIH Clinical Center"],
),
),
]
def create_mock_hypotheses() -> list[MechanismHypothesis]:
"""Create mock hypotheses for demonstration."""
return [
MechanismHypothesis(
drug="Metformin",
target="AMPK",
pathway="mTOR inhibition -> Autophagy activation",
effect="Clearance of tau and amyloid-beta aggregates",
confidence=0.85,
supporting_evidence=[
"https://pubmed.ncbi.nlm.nih.gov/19001324/",
"https://pubmed.ncbi.nlm.nih.gov/32109876/",
],
contradicting_evidence=[],
search_suggestions=[
"AMPK autophagy neurodegeneration",
"metformin tau clearance",
],
),
MechanismHypothesis(
drug="Metformin",
target="Glucose metabolism",
pathway="Improved neuronal energy homeostasis",
effect="Reduced oxidative stress and neuroinflammation",
confidence=0.70,
supporting_evidence=["https://pubmed.ncbi.nlm.nih.gov/31256789/"],
contradicting_evidence=[],
search_suggestions=[
"metformin brain glucose metabolism",
"neuronal insulin resistance alzheimer",
],
),
]
async def run_mock_demo() -> None:
"""Run full pipeline with mock data (no API keys needed)."""
print_header("DeepCritical Full Stack Demo (MOCK MODE)")
print("Running with synthetic data - no API keys required.\n")
evidence = create_mock_evidence()
hypotheses = create_mock_hypotheses()
# Step 1: Show evidence
print_step(1, "SEARCH (Phase 2) - Evidence Collection")
print(f"Collected {len(evidence)} pieces of evidence:\n")
for i, e in enumerate(evidence, 1):
print(f" [{i}] {e.citation.source.upper()}: {e.citation.title[:50]}...")
print(f" {e.content[:80]}...")
print()
# Step 2: Embedding deduplication
print_step(2, "EMBEDDINGS (Phase 6) - Semantic Deduplication")
try:
from src.services.embeddings import EmbeddingService
service = EmbeddingService()
unique = await service.deduplicate(evidence, threshold=0.85)
print(f"Original: {len(evidence)} papers")
print(f"After deduplication: {len(unique)} unique papers")
print("(Semantic duplicates removed by meaning, not just URL)")
except ImportError:
print("Embedding dependencies not installed - skipping deduplication")
unique = evidence
# Step 3: Hypothesis generation
print_step(3, "HYPOTHESIS (Phase 7) - Mechanistic Reasoning")
print(f"Generated {len(hypotheses)} hypotheses:\n")
for i, h in enumerate(hypotheses, 1):
print(f" Hypothesis {i} (Confidence: {h.confidence:.0%})")
print(f" {h.drug} -> {h.target} -> {h.pathway} -> {h.effect}")
print(f" Suggested searches: {', '.join(h.search_suggestions)}")
print()
# Step 4: Judge assessment
print_step(4, "JUDGE (Phase 3) - Evidence Assessment")
print("Assessment Results:")
print(" Mechanism Score: 8/10 (Strong mechanistic evidence)")
print(" Clinical Score: 7/10 (Phase 2 trial + observational data)")
print(" Confidence: 75%")
print(" Recommendation: SYNTHESIZE (Evidence sufficient)")
print()
# Step 5: Report generation
print_step(5, "REPORT (Phase 8) - Structured Scientific Report")
report = f"""
# Drug Repurposing Analysis: Metformin for Alzheimer's Disease
## Executive Summary
This analysis evaluated metformin as a potential therapeutic for Alzheimer's
disease. Evidence from {len(unique)} sources supports a plausible mechanism
through AMPK activation and mTOR inhibition, leading to enhanced autophagy
and clearance of pathological protein aggregates. Clinical data shows
promising risk reduction in observational studies and early trial results.
## Research Question
Can metformin, a type 2 diabetes medication, be repurposed for the prevention
or treatment of Alzheimer's disease?
## Methodology
- Searched PubMed and web sources for "metformin Alzheimer's disease"
- Applied semantic deduplication to remove redundant findings
- Generated mechanistic hypotheses using LLM reasoning
- Evaluated evidence quality with structured assessment
## Hypotheses Tested
- **Metformin -> AMPK -> mTOR inhibition -> Neuroprotection** (SUPPORTED)
- 4 supporting papers, 0 contradicting
- **Metformin -> Glucose metabolism -> Reduced oxidative stress** (PARTIAL)
- 2 supporting papers, requires more investigation
## Mechanistic Findings
Strong evidence supports AMPK activation as the primary mechanism. Metformin
crosses the blood-brain barrier and achieves therapeutic concentrations in
hippocampus and cortex. Downstream effects include:
- mTOR inhibition
- Autophagy activation
- Tau dephosphorylation
- Amyloid-beta clearance
## Clinical Findings
- Observational: 35% risk reduction (HR=0.65, n=100,000)
- Preclinical: 45% reduction in tau phosphorylation in AD mice
- Phase 2 trial: 18% slower cognitive decline vs placebo
## Drug Candidates
- **Metformin** - Primary candidate with established safety profile
## Limitations
- Abstract-level analysis only
- Observational data subject to confounding
- Limited RCT data available
- Optimal dosing for neuroprotection unclear
## Conclusion
Metformin shows strong potential for Alzheimer's disease prevention/treatment.
The AMPK-mTOR-autophagy mechanism is well-supported. Recommend Phase 3 trials
with cognitive endpoints.
## References
"""
max_authors_display = 2
for i, e in enumerate(unique[:6], 1):
authors = ", ".join(e.citation.authors[:max_authors_display])
if len(e.citation.authors) > max_authors_display:
authors += " et al."
ref_line = (
f"{i}. {authors}. *{e.citation.title}*. "
f"{e.citation.source.upper()} ({e.citation.date}). "
f"[Link]({e.citation.url})"
)
report += ref_line + "\n"
report += f"""
---
*Report generated from {len(unique)} papers across 3 search iterations.
Confidence: 75%*
"""
print(report)
async def _run_search_iteration(
query: str,
iteration: int,
evidence_store: dict[str, Any],
all_evidence: list[Evidence],
search_handler: Any,
embedding_service: Any,
) -> list[Evidence]:
"""Run a single search iteration with deduplication."""
search_queries = [query]
if evidence_store.get("hypotheses"):
for h in evidence_store["hypotheses"][-2:]:
search_queries.extend(h.search_suggestions[:1])
for q in search_queries[:2]:
result = await search_handler.execute(q, max_results_per_tool=5)
print(f" '{q}' -> {result.total_found} results")
new_unique = await embedding_service.deduplicate(result.evidence)
print(f" After dedup: {len(new_unique)} unique")
all_evidence.extend(new_unique)
evidence_store["current"] = all_evidence
evidence_store["iteration_count"] = iteration
return all_evidence
async def run_real_demo(query: str, max_iterations: int) -> None:
"""Run full pipeline with real API calls."""
print_header("DeepCritical Full Stack Demo")
print(f"Query: {query}")
print(f"Max iterations: {max_iterations}")
print("Mode: REAL (Live API calls)\n")
# Import real components
from src.agent_factory.judges import JudgeHandler
from src.agents.hypothesis_agent import HypothesisAgent
from src.agents.report_agent import ReportAgent
from src.services.embeddings import EmbeddingService
from src.tools.pubmed import PubMedTool
from src.tools.search_handler import SearchHandler
from src.tools.websearch import WebTool
# Initialize services
embedding_service = EmbeddingService()
search_handler = SearchHandler(tools=[PubMedTool(), WebTool()], timeout=30.0)
judge_handler = JudgeHandler()
# Shared evidence store
evidence_store: dict[str, Any] = {"current": [], "hypotheses": [], "iteration_count": 0}
all_evidence: list[Evidence] = []
for iteration in range(1, max_iterations + 1):
print_step(iteration, f"ITERATION {iteration}/{max_iterations}")
# Step 1: Search
print("\n[Search] Querying PubMed and Web...")
all_evidence = await _run_search_iteration(
query, iteration, evidence_store, all_evidence, search_handler, embedding_service
)
# Step 2: Generate hypotheses (first iteration only)
if iteration == 1:
print("\n[Hypothesis] Generating mechanistic hypotheses...")
hypothesis_agent = HypothesisAgent(evidence_store, embedding_service)
hyp_response = await hypothesis_agent.run(query)
print(hyp_response.messages[0].text[:500] + "...")
# Step 3: Judge
print("\n[Judge] Assessing evidence quality...")
assessment = await judge_handler.assess(query, all_evidence)
print(f" Mechanism: {assessment.details.mechanism_score}/10")
print(f" Clinical: {assessment.details.clinical_evidence_score}/10")
print(f" Recommendation: {assessment.recommendation}")
if assessment.recommendation == "synthesize":
print("\n[Judge says] Evidence sufficient! Generating report...")
evidence_store["last_assessment"] = assessment.details.model_dump()
break
next_queries = assessment.next_search_queries[:2]
print(f"\n[Judge says] Need more evidence. Next queries: {next_queries}")
query = assessment.next_search_queries[0] if assessment.next_search_queries else query
# Step 4: Generate report
print_step(iteration + 1, "REPORT GENERATION")
report_agent = ReportAgent(evidence_store, embedding_service)
report_response = await report_agent.run(query)
print("\n" + "=" * 70)
print("FINAL RESEARCH REPORT")
print("=" * 70)
print(report_response.messages[0].text)
async def main() -> None:
"""Entry point."""
parser = argparse.ArgumentParser(
description="DeepCritical Full Stack Demo (Phases 1-8)",
formatter_class=argparse.RawDescriptionHelpFormatter,
epilog="""
Examples:
# Mock mode (no API keys)
uv run python examples/full_stack_demo/run_full.py --mock
# Real mode with metformin query
uv run python examples/full_stack_demo/run_full.py "metformin alzheimer"
# Sildenafil for heart failure
uv run python examples/full_stack_demo/run_full.py "sildenafil heart failure" -i 3
""",
)
parser.add_argument(
"query",
nargs="?",
default="metformin Alzheimer's disease",
help="Research query",
)
parser.add_argument(
"--mock",
action="store_true",
help="Run with mock data (no API keys needed)",
)
parser.add_argument(
"-i",
"--iterations",
type=int,
default=2,
help="Max search iterations (default: 2)",
)
args = parser.parse_args()
if args.mock:
await run_mock_demo()
else:
# Check for API keys
if not (os.getenv("OPENAI_API_KEY") or os.getenv("ANTHROPIC_API_KEY")):
print("Error: Real mode requires OPENAI_API_KEY or ANTHROPIC_API_KEY")
print("Use --mock for demo without API keys.")
sys.exit(1)
await run_real_demo(args.query, args.iterations)
print("\n" + "=" * 70)
print(" DeepCritical Full Stack Demo Complete!")
print(" Phases demonstrated: Foundation -> Search -> Judge -> UI ->")
print(" Magentic -> Embeddings -> Hypothesis -> Report")
print("=" * 70 + "\n")
if __name__ == "__main__":
asyncio.run(main())
|