BioMedGPS Explainer - Drug Discovery Toolkit

Key Features

Drug Prediction

Predict potential drugs using Knowledge Graph Embedding (KGE) models with advanced algorithms

Network Analysis

Comprehensive drug-disease-gene pathway analysis with centrality calculations and PPI network analysis

Pathway Enrichment

Advanced pathway overlap analysis between drugs and diseases for deeper insights

Shared Annotations

Statistical analysis of shared genes and diseases between drugs and diseases

Smart Filtering

Advanced drug filtering with support for complex logical expressions

Visualization

Automatic chart generation with interactive plots and comprehensive HTML reports

Quick Start

Installation

git clone <repository-url>
cd biomedgps-explainer
pip install -e .

Model Setup

Models are automatically downloaded from Weights & Biases (wandb). Browse available models at wandb.ai/yjcyxky/biomedgps-kge-v1 to find different model run IDs.

# No manual preparation required!
# Use --model-run-id to specify which model to download

Run Analysis

python3 examples/run_full_example.py
# Or use CLI
biomedgps-explainer run --disease-id MONDO:0004979 --model-run-id 6vlvgvfq --output-dir results/

View Results

Check the results/ directory for comprehensive analysis reports and visualizations

Live Demo

Asthma Drug Discovery Analysis

Complete analysis report with interactive visualizations

View Report

Documentation

User Guide

Complete step-by-step guide for using BioMedGPS Explainer

Read Guide

API Reference

Comprehensive API documentation with examples

View API

Examples

Real-world examples and use cases

View Examples

Model Usage

Detailed guide for working with KGE models

Learn More