Comparing phenotypic manifolds with Kompot: Detecting differential abundance and gene expression at single-cell...

Area of Science:

• Computational Biology
• Genomics
• Immunology

Background:

• Single-cell technologies generate complex, high-dimensional data.
• Comparing cellular states and gene expression across multiple conditions is challenging.
• Existing methods often lack the ability to capture nuanced, localized changes.

Purpose of the Study:

• To introduce Kompot, a statistical framework for holistic comparison of multi-condition single-cell datasets.
• To enable simultaneous analysis of differential abundance and differential expression.
• To provide a method for understanding condition-specific effects at single-cell resolution.

Main Methods:

• Kompot models cell distributions and gene expression as continuous functions over low-dimensional cell state representations.
• It supports inference of differential abundance and differential expression with calibrated uncertainty.
• The framework integrates global and cell-state-specific analyses.

Main Results:

• Applied to aging murine bone marrow, Kompot identified a continuum of shifts in hematopoietic stem and mature cell states.
• Transcriptional remodeling of monocytes was observed independently of compositional changes.
• Divergent regulation of oxidative stress response genes across cell types was revealed.

Conclusions:

• Kompot effectively captures both global and cell-state-specific effects of perturbations in single-cell data.
• The framework elucidates how aging reshapes cellular identity and regulatory programs in the hematopoietic system.
• Kompot is broadly applicable to dissecting condition-specific effects in complex biological landscapes.