Improving generalization of machine learning-identified biomarkers with causal modeling: an investigation into immune receptor diagnostics
This work addresses the challenge of developing clinically applicable diagnostics from high-dimensional data, but it is incremental as it builds on existing biomarker methods with a causal perspective.
The paper tackles the problem of improving generalization of machine learning-identified biomarkers by using causal modeling, specifically applied to adaptive immune receptor repertoires, and argues that this approach enhances robustness by identifying stable relationships and guiding adjustments for population variations.
Machine learning is increasingly used to discover diagnostic and prognostic biomarkers from high-dimensional molecular data. However, a variety of factors related to experimental design may affect the ability to learn generalizable and clinically applicable diagnostics. Here, we argue that a causal perspective improves the identification of these challenges and formalizes their relation to the robustness and generalization of machine learning-based diagnostics. To make for a concrete discussion, we focus on a specific, recently established high-dimensional biomarker - adaptive immune receptor repertoires (AIRRs). Through simulations, we illustrate how major biological and experimental factors of the AIRR domain may influence the learned biomarkers. In conclusion, we argue that causal modeling improves machine learning-based biomarker robustness by identifying stable relations between variables and by guiding the adjustment of the relations and variables that vary between populations.