Single-cell entropy for accurate estimation of differentiation potency from a cell's transcriptome.

Area of Science:

• Single-cell biology
• Systems biology
• Bioinformatics

Background:

• Quantifying single-cell differentiation potential is crucial for understanding development and disease.
• Existing methods for assessing cell potency have limitations.

Purpose of the Study:

• To develop a novel, accurate, and robust method for estimating single-cell differentiation potency.
• To identify normal and cancer stem cell phenotypes using computational approaches.

Main Methods:

• Analysis of over 7,000 single-cell RNA sequencing profiles.
• Computation of signalling entropy (transcriptome signalling promiscuity) within an interaction network.
• Comparison of signalling entropy with other entropy-based measures.

Main Results:

• Signalling entropy accurately approximates single-cell differentiation potency without feature selection.
• This method outperforms other entropy-based potency estimates.
• Signalling entropy identifies known cell subpopulations and drug-resistant cancer stem cells, including those from circulating tumor cells.
• Expression heterogeneity within single-cell populations is shown to be regulated.

Conclusions:

• Signalling entropy offers a powerful in silico tool for estimating differentiation potency and plasticity in single cells and bulk samples.
• This approach facilitates the identification of normal and cancer stem cell phenotypes.