Inferring Causal Gene Regulatory Networks from Coupled Single-Cell Expression Dynamics Using Scribe.

Area of Science:

• Genomics
• Systems Biology
• Computational Biology

Background:

• Understanding gene regulatory networks is crucial for deciphering cellular mechanisms.
• Single-cell experiments offer unprecedented resolution but pose challenges for network reconstruction.

Purpose of the Study:

• To introduce Scribe, a toolkit for detecting causal gene regulatory interactions.
• To evaluate the utility of single-cell data for network reconstruction.
• To identify requirements for accurate causal inference in gene regulation.

Main Methods:

• Development of the Scribe toolkit utilizing restricted directed information.
• Application of Scribe and comparative analysis with other causal network reconstruction methods.
• Evaluation using diverse single-cell measurement types, including pseudotime and true time-series data.

Main Results:

• Scribe effectively detects and visualizes causal regulatory interactions.
• Performance of causal network reconstruction significantly drops with pseudotime-ordered single-cell data compared to true time-series data.
• Temporal coupling between measurements is essential for robust causal inference.

Conclusions:

• Accurate causal inference in gene regulatory networks from single-cell data necessitates temporal coupling.
• Existing methods may be limited by the temporal resolution of single-cell experiments.
• RNA velocity and similar methods can partially restore temporal coupling for improved analysis.