Considerations of automated machine learning in clinical metabolic profiling: Altered homocysteine plasma concentration associated with metformin exposure

TL;DR

This study evaluates the use of AutoML, specifically TPOT, in clinical metabolic profiling, highlighting its potential and challenges in identifying metabolite associations like homocysteine levels linked to metformin exposure.

Contribution

The paper introduces a tandem rank-accuracy measure for feature selection and demonstrates methods to adjust AutoML for clinical confounders in metabolic profiling.

Findings

AutoML with TPOT can identify novel metabolite associations.

Adjustment for clinical confounders improves AutoML sensitivity.

Long-term metformin exposure linked to increased homocysteine levels.

Abstract

With the maturation of metabolomics science and proliferation of biobanks, clinical metabolic profiling is an increasingly opportunistic frontier for advancing translational clinical research. Automated Machine Learning (AutoML) approaches provide exciting opportunity to guide feature selection in agnostic metabolic profiling endeavors, where potentially thousands of independent data points must be evaluated. In previous research, AutoML using high-dimensional data of varying types has been demonstrably robust, outperforming traditional approaches. However, considerations for application in clinical metabolic profiling remain to be evaluated. Particularly, regarding the robustness of AutoML to identify and adjust for common clinical confounders. In this study, we present a focused case study regarding AutoML considerations for using the Tree-Based Optimization Tool (TPOT) in metabolic profiling of exposure to metformin in a biobank cohort. First, we propose a tandem rank-accuracy measure to guide agnostic feature selection and corresponding threshold determination in clinical metabolic profiling endeavors. Second, while AutoML, using default parameters, demonstrated potential to lack sensitivity to low-effect confounding clinical covariates, we demonstrated residual training and adjustment of metabolite features as an easily applicable approach to ensure AutoML adjustment for potential confounding characteristics. Finally, we present increased homocysteine with long-term exposure to metformin as a potentially novel, non-replicated metabolite association suggested by TPOT; an association not identified in parallel clinical metabolic profiling endeavors. While considerations are recommended, including adjustment approaches for clinical confounders, AutoML presents an exciting tool to enhance clinical metabolic profiling and advance translational research endeavors.