SCOUP: a probabilistic model based on the Ornstein-Uhlenbeck process to analyze single-cell expression data during...

Area of Science:

• Genomics
• Computational Biology
• Systems Biology

Background:

• Single-cell technologies enable comprehensive cellular state quantification, crucial for understanding cell differentiation.
• Existing methods for analyzing single-cell expression data have limitations, particularly in modeling differentiation processes.

Purpose of the Study:

• To introduce SCOUP, a novel method for elucidating cell differentiation dynamics.
• To improve the accuracy of pseudo-time and cell lineage estimation in single-cell expression data analysis.

Main Methods:

• SCOUP utilizes a stochastic process-based approach to directly model gene expression dynamics during differentiation.
• The method estimates the degree of differentiation (pseudo-time) and cell fate for individual cells.
• A novel correlation calculation method is proposed for downstream analysis of regulatory relationships.

Main Results:

• SCOUP demonstrates superior performance in pseudo-time estimation compared to previous methods, especially for single-cell RNA-seq data.
• Cell lineage inference is more accurate with SCOUP, particularly for cells at early bifurcation stages.
• Application of the novel correlation method identified a potential key regulator and novel clusters in gene regulatory networks.

Conclusions:

• SCOUP offers a robust stochastic process-based framework for analyzing single-cell expression data during differentiation.
• The method provides enhanced accuracy in pseudo-time and cell lineage estimation.
• SCOUP facilitates downstream analyses, including regulatory network inference, and is publicly available.