A phenotype-driven and evidence-governed framework for knowledge graph enrichment and hypotheses discovery in population data

confirmatory KG construction: Current knowledge graph construction methods focus on recovering known relationships instead of identifying novel or context-dependent nodes.

trivial or redundant KG expansion: Knowledge graph expansion risks including trivial or redundant knowledge when balancing confirmation and discovery is not controlled.

LLM hallucination in grounded outputs: Retrieval-augmented LLM settings require grounding to reduce hallucination rates.

trivial or redundant KG expansion: Knowledge graph expansion risks including trivial or redundant knowledge when balancing confirmation and discovery is not controlled.

LLM hallucination in grounded outputs: Retrieval-augmented LLM settings require grounding to reduce hallucination rates.

phenotype-driven evidence-governed KG framework: The paper proposes a phenotype-driven and evidence-governed framework for structured hypothesis discovery and controlled knowledge graph expansion.

GNN causal probabilistic LLM pipeline: The framework integrates GNNs, causal inference, probabilistic reasoning, and LLMs for phenotype discovery, hypothesis generation, and claim extraction.

multi-objective Pareto KG expansion: The approach formulates KG expansion as multi-objective optimization and uses Pareto-optimal selection to identify non-dominated claims balancing relevance, validation, and novelty.

phenotype-driven evidence-governed KG framework: The paper proposes a phenotype-driven and evidence-governed framework for structured hypothesis discovery and controlled knowledge graph expansion.

GNN causal probabilistic LLM pipeline: The framework integrates GNNs, causal inference, probabilistic reasoning, and LLMs for phenotype discovery, hypothesis generation, and claim extraction.

more interpretable phenotypes:

context-dependent causal structures revealed:

high-quality evidence-aligned claims: