scGRN-Entropy: Inferring cell differentiation trajectories using single-cell data and gene regulation network-based...

Area of Science:

• Computational Biology
• Genomics
• Systems Biology

Background:

• Cell differentiation research is crucial for understanding life processes and diseases like cancer.
• Current methods for inferring cell differentiation trajectories from single-cell RNA sequencing (scRNA-seq) data rely on static gene expression, limiting accuracy.
• Advancements in precision medicine and therapeutics depend on accurate cell trajectory inference.

Purpose of the Study:

• To introduce scGRN-Entropy, a novel method for inferring cell differentiation trajectories and pseudotime from scRNA-seq data.
• To improve the accuracy of cell differentiation trajectory inference by incorporating dynamic gene regulatory network (GRN) information.
• To provide a more robust tool for analyzing cellular processes and disease mechanisms.

Main Methods:

• Constructing an undirected graph integrating static gene expression and dynamic GRN relationships.
• Refining graph edges using pseudotime inferred from cell entropy within the GRN space.
• Applying the Minimum Spanning Tree (MST) algorithm to derive the final cell differentiation trajectory.

Main Results:

• scGRN-Entropy demonstrates superior performance in inferring cell differentiation trajectories.
• Validation on eight diverse scRNA-seq datasets confirms the method's accuracy and robustness.
• Comparative analysis shows improved results over existing state-of-the-art methods.

Conclusions:

• scGRN-Entropy offers a significant advancement in analyzing cell differentiation from scRNA-seq data.
• Incorporating dynamic GRN information enhances the accuracy of trajectory inference.
• This method has broad implications for understanding developmental biology and disease pathogenesis.