Comprehensive benchmarking of RNA velocity methods across single-cell datasets

Area of Science:

• Computational Biology
• Single-Cell Genomics
• Systems Biology

Background:

• RNA velocity analysis is crucial for understanding cellular dynamics from single-cell RNA sequencing data.
• The growing number of computational tools necessitates systematic evaluation.
• Existing comparisons lack comprehensive scope and standardized metrics, hindering user guidance.

Purpose of the Study:

• To conduct a comprehensive benchmark of computational RNA velocity tools.
• To evaluate methods across diverse biological and technical scenarios using context-specific metrics.
• To provide task-aware guidance for selecting appropriate RNA velocity methods.

Main Methods:

• Systematic evaluation of 19 RNA velocity tools, encompassing 30 distinct methods.
• Assessment of 25 RNA-only methods across core tasks: directional consistency, temporal precision, negative control robustness, and sequencing depth stability.
• Evaluation of 5 multimodal-enhanced methods on multimodal integration using 34 datasets from 26 real-world and 8 simulated scenarios.

Main Results:

• Identified a trade-off between directional consistency and negative control robustness.
• Observed distinct group behaviors based on temporal modeling strategies.
• Highlighted variability influenced by sequencing depth and quantification choices.
• Revealed methodological gaps in gene dependence modeling, temporal inference, and multimodal architectures.

Conclusions:

• Established a unified framework for evaluating RNA velocity methods.
• Provided task-aware guidance to aid method selection based on specific biological contexts and technical constraints.
• Emphasized context-specific selection over a single overall ranking.