Uncertainty Quantification for Atlas-Level Cell Type Transfer

TL;DR

This paper introduces uncertainty quantification methods for cell type classification in large-scale single-cell atlases, improving robustness and detection of unseen cell types amidst domain shifts.

Contribution

It is the first to apply and benchmark uncertainty quantification techniques for cell type transfer in single-cell atlases, highlighting their advantages over existing models.

Findings

Current models lack calibration and robustness.

Uncertainty-aware models better detect unseen cell types.

Uncertainty estimates improve interpretation of cell type predictions.

Abstract

Single-cell reference atlases are large-scale, cell-level maps that capture cellular heterogeneity within an organ using single cell genomics. Given their size and cellular diversity, these atlases serve as high-quality training data for the transfer of cell type labels to new datasets. Such label transfer, however, must be robust to domain shifts in gene expression due to measurement technique, lab specifics and more general batch effects. This requires methods that provide uncertainty estimates on the cell type predictions to ensure correct interpretation. Here, for the first time, we introduce uncertainty quantification methods for cell type classification on single-cell reference atlases. We benchmark four model classes and show that currently used models lack calibration, robustness, and actionable uncertainty scores. Furthermore, we demonstrate how models that quantify uncertainty are better suited to detect unseen cell types in the setting of atlas-level cell type transfer.