Entropy-based inference of transition states and cellular trajectory for single-cell transcriptomics

Area of Science:

• Computational Biology
• Genomics
• Bioinformatics

Background:

• Single-cell RNA sequencing (scRNA-seq) enables high-resolution analysis of cellular processes.
• Inferring pseudo-time trajectories is crucial for understanding dynamic biological events.
• Identifying transitional cell states remains a challenge in trajectory inference.

Purpose of the Study:

• To introduce scTite, a new method for single-cell trajectory inference.
• To accurately identify transitional cell states and reconstruct complex cell trajectories from scRNA-seq data.
• To improve the accuracy of pseudo-time trajectory inference.

Main Methods:

• Developed a novel metric, transition entropy, to quantify cell uncertainty across clusters.
• Employed Wasserstein distance for inter-cluster analysis and minimum spanning tree construction.
• Integrated signaling entropy and partial correlation for identifying transition paths and refining trajectories.

Main Results:

• scTite successfully identified cell states and transition cells.
• The method demonstrated superior accuracy in reconstructing cell trajectories compared to existing algorithms.
• Performance was validated on multiple real-world scRNA-seq datasets.

Conclusions:

• scTite provides a robust and accurate approach for single-cell trajectory inference.
• The method effectively handles complex biological processes with intermediate cell states.
• scTite offers a valuable tool for researchers analyzing dynamic cellular systems.