Bayesian Variable Selection on Small Sample Trial Data via Adaptive Posterior-Informed Shrinkage Prior

2025-11-17

AI Summary

In cell and gene therapy and ultra-rare disease research, variable selection is severely challenged by extremely small sample sizes (e.g., n < 20). To address this, we propose an adaptive posterior information-shrinkage prior that integrates a data-driven external information borrowing mechanism, a hybrid shrinkage prior structure, and an empirical null distribution–based decision rule within a Bayesian variable selection framework. This enables robust incorporation of historical control data while mitigating overfitting and instability. Simulation studies demonstrate substantial improvements in true positive rate and selection stability under ultra-small samples compared to conventional methods. Applied to the CIT06 clinical trial, the method successfully identified biomarkers significantly associated with C-peptide levels—validating its practical efficacy and generalizability. The approach advances Bayesian variable selection for settings where traditional frequentist or standard Bayesian methods fail due to insufficient data.

Technology Category

• Machine Learning: Bayesian Learning
• Search and Optimization: Sampling/Simulation-based Search
• Reasoning under Uncertainty: Stochastic Optimization

Application Category

• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• User Modeling, Personalization and Recommendation: User privacy protection in personalized systems
• Economics, Online Markets and Human Computation: Data quality aspects of human-annotated datasets

Abstract

Identifying variables associated with clinical endpoints is of much interest in clinical trials. With the rapid growth of cell and gene therapy (CGT) and therapeutics for ultra-rare diseases, there is an urgent need for statistical methods that can detect meaningful associations under severe sample-size constraints. Motivated by data-borrowing strategies for historical controls, we propose the Adaptive Posterior-Informed Shrinkage Prior (APSP), a Bayesian approach that adaptively borrows information from external sources to improve variable-selection efficiency while preserving robustness across plausible scenarios. APSP builds upon existing Bayesian data borrowing frameworks, incorporating data-driven adaptive information selection, structure of mixture shrinkage informative priors and decision making with empirical null to enhance variable selection performances under small sample size. Extensive simulations show that APSP attains better efficiency relative to traditional and popular data-borrowing and Bayesian variable-selection methods while maintaining robustness under linear relationships. We further applied APSP to identify variables associated with peak C-peptide at Day 75 from the Clinical Islet Transplantation (CIT) Consortium study CIT06 by borrowing information from the study CIT07.

Problem

• Identifying clinical endpoint variables under severe sample-size constraints
• Developing Bayesian methods for small sample cell and gene therapy trials
• Improving variable selection efficiency through adaptive external data borrowing

Innovation

• Adaptive Posterior-Informed Shrinkage Prior for Bayesian analysis
• Borrows information from external sources adaptively
• Enhances variable selection under small sample sizes