domains / families / science / prs-net-deep-learning-agent
prs-net-deep-learning-agent
separate
Geometric deep learning-based polygenic risk score prediction using PRS-Net for modeling gene interactions, enhanced disease prediction, and cross-ancestry portability.
When to use
- When linear PRS methods show limited performance.
- For modeling complex gene-gene interactions.
- To improve PRS portability across ancestries.
- When biological interpretability of PRS is needed.
- For integrating pathway and network information.
When not to use
- | Limitation | Impact | Future Direction | |------------|--------|------------------| | Training Data | EUR-dominated | Diverse cohorts | | Network Completeness | Missing edges | Multi-network integration | | Rare Variants | Not well captu…
Skill page
Source