Gene & RNA Therapeutics Intelligence Pipeline

Leverage AI and Literature For Insights

Project Overview

AI‑powered PubMed for BD, R&D, and Competitive Strategy This project builds an end‑to‑end intelligence pipeline that automatically retrieves, structures, analyzes, and visualizes scientific and patent evidence for gene therapy and RNA therapeutics. It is designed to support BD&L, competitive intelligence, R&D strategy, and due diligence workflows in pharma and biotech. The system integrates PubMed,LLM‑based structured extraction, evidence summaries, competitive landscape analysis, and BD‑ready Word report generation—all in one automated workflow.

Dataset

Pubmed

Key Capabilities

1. Automated Literature:

• Queries PubMed using NCBI Entrez
• Supports configurable company, modality, and disease focus
• Extracts abstracts and metadata for downstream analysis

2. LLM‑Powered Structured Evidence Extraction Strict JSON extraction (sample size, study design, phase, endpoints, efficacy, safety):

• Includes source sentences for auditability
• Validates completeness before downstream use

3. LLM Summaries & Competitive Intelligence Generates 3–4 sentence evidence summaries:

• Produces competitive landscape analysis
• Key competitors
• MoA comparisons
• Efficacy & safety positioning
• Strategic threat level

4. Visual Intelligence (Charts) Two branded charts are automatically generated:

• Clinical Phase Distribution (Donut Chart)
• Gene vs RNA Therapy Strategic Scorecard (Horizontal Bar Chart)

5. BD‑Ready Word Report Generation Automatically produces a polished .docx report containing:

• Title page with metadata
• Visual intelligence charts
• Structured evidence tables
• LLM summaries
• Competitive analysis
• Exported to outputs/<drug_name>_BD_READY_REPORT.docx

Example Use Case

The included configuration demonstrates how the pipeline can answer strategic questions for gene & RNA therapeutics portfolio:

• What scientific and patent evidence exists for AbbVie’s gene/RNA programs?
• What is the clinical phase distribution across modalities?
• How do gene therapy and RNA therapy compare across BD‑critical metrics?
• What are the key competitors and their MoA?
• What are the safety/efficacy signals across studies?
• Where are the white‑space opportunities?

Project Structure

├── main.py # Main pipeline entry point

├── outputs/ # Auto-generated charts, CSVs, and Word reports

├── requirements.txt # Python dependencies

└── README.md # Project documentation

Methodology

• Step 1 — Configure the drug/company focus
• Step 2 — Retrieve PubMed & Patent Data
• Step 3 — Extract Structured Evidence
• Step 4 — Summaries & Competitive Analysis
• Step 5 — Visual Intelligence
• Step 6 — Export BD‑Ready Report

Tech Stack

Python

Example Outputs

1. Clinical Phase Distribution (Donut Chart) Shows aggregated development stage distribution across extracted records + supplemental pipeline data.

2. Therapeutic Modality Scorecard Compares gene therapy vs RNA therapy across: Delivery maturity Tissue accessibility Immune risk Commercial potential Clinical readiness Competitive moat

3. BD‑Ready Word Report

Tech Stack

NCBI Entrez API Groq + Llama 3.1 Data processing Python Pandas Matplotlib export python-docx

Running the Pipeline

1. Install dependencies pip install -r requirements.txt

2. Set environment variables NCBI_EMAIL=your_email NCBI_API_KEY=your_ncbi_key GROQ_API_KEY=your_groq_key LENS_API_KEY=your_lens_key

3. Run python main.py

4. Find outputs All generated files will appear in: outputs/