Modeling clinical outcome using multiple correlated functional biomarkers: A Bayesian approach.

Area of Science:

• Biostatistics
• Biomedical Data Analysis
• Statistical Modeling

Background:

• Biomedical studies often measure functional biomarkers over time or space, with inherent measurement error.
• Evaluating associations between clinical endpoints and multiple functional biomarkers is crucial but challenging due to complex correlation structures.

Purpose of the Study:

• To propose a novel Bayesian approach for modeling clinical outcomes with multiple functional biomarkers.
• To address the limitation of existing methods that do not account for correlations between multiple functional biomarkers.

Main Methods:

• Development of a Bayesian statistical model to jointly analyze multiple functional biomarkers and a clinical outcome.
• Simulation studies to evaluate the performance of the proposed method under various correlation settings.
• Application of the method to real-world data from a colorectal cancer risk study.

Main Results:

• The proposed Bayesian approach demonstrates good performance in finite samples.
• The method outperforms existing approaches when moderate to substantial correlation is present between biomarkers.
• Identified significant associations between colorectal cancer risk and specific functional biomarkers (APC, TGF-α) in the colorectal crypts.

Conclusions:

• The novel Bayesian approach effectively models associations between clinical endpoints and multiple correlated functional biomarkers.
• Accounting for inter-biomarker correlation improves the accuracy of risk prediction models.
• The findings highlight the role of APC and TGF-α in colorectal cancer development within specific cellular regions.