SCLC Cannabinoid Literature Search - Quick Start Guide

Installation

1. Install required packages:

pip install requests pandas numpy matplotlib seaborn scikit-learn networkx openpyxl

2. Configure your PubMed API Key (optional but recommended):
   • Go to: https://www.ncbi.nlm.nih.gov/account/
   • Sign in or create an account
   • Go to API Key Management
   • Generate your API key
   • Edit config.py and replace YOUR_PUBMED_API_KEY_HERE with your actual key

Running the Search

Basic usage (will use key from config.py if configured):

python sclc_complete_search.py

Override with different PubMed API key:

python sclc_complete_search.py --pubmed-key YOUR_KEY_HERE

Advanced options:

python sclc_complete_search.py --pubmed-key YOUR_KEY --max-per-query 50 --top-papers 200

What the Script Does

1. Phase 1: Mechanism-Based Search

   • Searches 64 carefully crafted queries across 8 mechanism categories
   • Uses PubMed, Europe PMC, and Semantic Scholar simultaneously
   • Focuses on SCLC-specific vulnerabilities identified by Gemini

2. Phase 2: Deduplication

   • Removes duplicate papers found across multiple searches
   • Preserves the version with most complete information

3. Phase 3: Relevance Scoring

   • Calculates multi-factor relevance scores considering:
     • Mechanism novelty (SCLC-specific mechanisms scored higher)
     • Query specificity (more specific searches ranked higher)
     • Citation impact and influence
     • Recency and open access availability

4. Phase 4: Analysis Report

   • Generates comprehensive JSON report with:
     • Coverage by mechanism category
     • Top cited papers in each category
     • Research gaps identified
     • Search effectiveness metrics

5. Phase 5: Reading List

   • Creates prioritized reading list (CSV and Excel)
   • Ensures diversity across mechanism categories
   • Marks papers as High/Medium/Low priority

Output Files

All results are saved in a timestamped directory:

sclc_search_results_YYYYMMDD_HHMMSS/
├── sclc_raw_results_*.csv          # All search results
├── sclc_dedup_results_*.csv        # Deduplicated results  
├── sclc_analysis_report_*.json     # Comprehensive analysis
├── sclc_reading_list_*.csv         # Prioritized papers (CSV)
├── sclc_reading_list_*.xlsx        # Prioritized papers (Excel)
└── sclc_search.log                 # Detailed execution log

Mechanism Categories Searched

1. Direct Receptors - CB1/2, TRP channels, PPARs, orphan GPCRs
2. Cell Death - Apoptosis, autophagy, ferroptosis, necroptosis
3. SCLC Vulnerabilities - Neuroendocrine features, transcriptional subtypes, DLL3
4. Signaling Pathways - PI3K/AKT/mTOR, NF-κB, MAPK, JAK/STAT
5. Metabolic - Mitochondria, glycolysis, oxidative stress
6. Tumor Microenvironment - Angiogenesis, immune modulation, CAFs
7. Drug Resistance - ABC transporters, cancer stem cells, EMT
8. Novel Mechanisms - ER stress, ion channels, epigenetics

Interpreting Results

Relevance Scores (0-100):

• 80-100: Highly relevant, novel mechanisms, SCLC-specific
• 60-79: Good relevance, established pathways
• 40-59: Moderate relevance, may be tangential
• 0-39: Lower relevance, broad searches

Priority Categories:

• High: Read immediately - most likely to reveal new insights
• Medium: Important supporting literature
• Low: Background reading, established knowledge

Next Steps After Search

1. Review the analysis report (sclc_analysis_report_*.json)

   • Check coverage gaps
   • Identify high-potential mechanisms

2. Start with High priority papers in reading list

   • Focus on SCLC vulnerabilities and novel mechanisms
   • Look for papers that bridge cannabinoid effects with SCLC biology

3. Track promising mechanisms for experimental validation

   • Note specific assays mentioned in papers
   • Identify druggable targets

4. Consider follow-up searches for:

   • Specific compounds mentioned (e.g., WIN 55,212-2)
   • Mechanism combinations showing promise
   • Authors frequently appearing in results

Troubleshooting

Rate limit errors:

• The script includes delays between searches
• If you hit limits, wait 5 minutes and restart

No results for some queries:

• This is expected for novel mechanism combinations
• These represent research opportunities!

Memory issues with large searches:

• Reduce --max-per-query parameter
• Process in batches by mechanism category

Research Strategy Tips

1. Start with SCLC vulnerabilities - Highest novelty potential
2. Look for convergent mechanisms - Multiple papers pointing to same pathway
3. Note experimental methods - Build a toolkit for validation
4. Track citation networks - Use Semantic Scholar links to find related work
5. Identify combination opportunities - Mechanisms that could synergize

Contact and Updates

For questions or to report issues:

• Check the log file for detailed error messages
• Ensure all APIs are accessible from your network
• PubMed API key significantly improves reliability

Good luck with your research!

greentreetherapeutics