KWS Saat: From Graph to Gene: Knowledge Graphs and AI for Explainable Gene Discovery

This video presents a framework that uses knowledge graphs, link prediction, and generative AI to discover candidate genes for complex agricultural traits in an explainable way.

What it covers

Knowledge Graph Integration
The framework integrates genes, traits, and biological evidence into a unified knowledge graph.
Link Prediction
Link prediction is used to identify potential gene-trait associations within the knowledge graph.
Generative AI
Generative AI enhances hypothesis generation by exploring novel gene-trait relationships.
Explainable Hypotheses
The approach generates biologically meaningful and interpretable hypotheses, avoiding opaque AI predictions.
Agricultural Trait Discovery
The framework is applied to identify candidate genes linked to complex agricultural traits.