prs-net-deep-learning-agent - Deep Learning Polygenic Risk Score Prediction
Predicts disease risk using geometric deep learning on gene interaction networks
Tags
Updated: 2026-03-24Capabilities
Typical Inputs
Typical Outputs
What this skill does
- process genotypes
- summarize genes
- encode network
- predict risk
- quantify uncertainty
- identify pathways
- visualize modules
Inputs
- Genotypes (VCF/PLINK)
- PPI Network (GraphML)
- Gene Mapping (BED)
- Training Labels
- GWAS Summary
- Trained Model Weights
- STRING PPI Network
- BioGRID Network
- Reactome Network
- Gene Ontology
Outputs
- PRS-Net Score (CSV)
- Risk Percentile (CSV)
- Gene Importance (CSV)
- Pathway Enrichment (CSV)
- Module Visualization (PNG)
- Prediction Uncertainty (JSON)
Requirements
- Python 3.10+
- PyTorch
- PyTorch Geometric
- NetworkX
- igraph
- GPU recommended