InterFeat: An Automated Pipeline for Finding Interesting Hypotheses in Structured Biomedical Data

2025-05-18

AI Summary

This study addresses the challenge of automating the discovery of “interesting” feature–target relationships—including effect direction and underlying biological mechanisms—in structured biomedical data. We formally define “interestingness” as a unified metric integrating novelty, utility, and credibility. To this end, we propose an end-to-end hypothesis generation pipeline that synergistically integrates SHAP-based interpretable machine learning, knowledge graph reasoning, literature-based semantic retrieval, and large language model–assisted validation—enabling cross-disease, target-agnostic, and scalable hypothesis generation and ranking. Evaluated on eight disease cohorts from the UK Biobank, our method retrospectively rediscovered known risk factors years before their clinical recognition. Expert evaluation revealed that 40–53% of top-ranked hypotheses were deemed interesting—substantially outperforming baseline methods (0–7%). Moreover, 28% of 109 candidate hypotheses received endorsement from domain experts.

Technology Category

• Data Mining & Knowledge Management: Representing, Reasoning, and Using Provenance, Trust
• Knowledge Representation and Reasoning: Diagnosis and Abductive Reasoning
• Reasoning under Uncertainty: Causality

Application Category

• User Modeling, Personalization and Recommendation: Fairness-aware retrieval and ranking
• Semantics and Knowledge: Scalable techniques for the creation, curation, publication, maintenance, and consumption of large, Web-based, structured, reusable, knowledge graphs and ontologies
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking

Abstract

Finding interesting phenomena is the core of scientific discovery, but it is a manual, ill-defined concept. We present an integrative pipeline for automating the discovery of interesting simple hypotheses (feature-target relations with effect direction and a potential underlying mechanism) in structured biomedical data. The pipeline combines machine learning, knowledge graphs, literature search and Large Language Models. We formalize"interestingness"as a combination of novelty, utility and plausibility. On 8 major diseases from the UK Biobank, our pipeline consistently recovers risk factors years before their appearance in the literature. 40--53% of our top candidates were validated as interesting, compared to 0--7% for a SHAP-based baseline. Overall, 28% of 109 candidates were interesting to medical experts. The pipeline addresses the challenge of operationalizing"interestingness"scalably and for any target. We release data and code: https://github.com/LinialLab/InterFeat

Problem

• Automating discovery of interesting biomedical hypotheses from structured data
• Formalizing interestingness as novelty, utility, and plausibility
• Validating hypotheses years before literature appearance

Innovation

• Combines machine learning and knowledge graphs
• Uses literature search and Large Language Models
• Formalizes interestingness as novelty, utility, plausibility