Robust Inference of Cell-to-Cell Expression Variations from Single- and K-Cell Profiling

Area of Science:

• Single-cell biology
• Computational biology
• Genomics

Background:

• Measuring gene expression heterogeneity in single cells is crucial but challenging due to detection limits.
• Existing methods analyze single-cell or pooled (k-cell) data separately, limiting comprehensive analysis.
• Comparing cellular heterogeneity parameters (CHPs) across conditions can yield significant biological insights.

Purpose of the Study:

• To develop a Bayesian approach for inferring CHPs using single-cell, k-cell, or combined data.
• To enable comparison of CHPs within and across different cell populations or conditions.
• To improve the precision and reconciliation of CHP information from various data types.

Main Methods:

• A novel Bayesian framework was developed to infer CHPs.
• The approach accommodates single-cell data, k-cell pooled data, or a combination of both.
• Simulated and experimental data were used for validation and comparison.

Main Results:

• The integrated approach using both single- and k-cell data demonstrated improved precision and better reconciliation of CHP information.
• Each data type (single-cell or k-cell) showed advantages in specific scenarios.
• The method successfully identified CHP differences in inflammatory genes between resting and activated human macrophages.

Conclusions:

• The developed Bayesian approach provides a robust framework for assessing and comparing cellular heterogeneity.
• Integrating single-cell and k-cell data offers significant advantages for quantifying gene expression variability.
• This method enhances our understanding of biological variability in gene expression within and across conditions.