Researchers recently introduced veloAgent to advance our understanding of tissue organization and developmental biology. This innovative framework utilizes spatially-informed RNA velocity to estimate transcriptional kinetics with high accuracy. By integrating molecular data with spatial cues, the tool provides a more complete picture of how cells transition between states within their local microenvironments. This approach bridges the gap between static gene expression snapshots and dynamic cellular processes.

Current methods for analyzing single-cell transcriptomics often struggle with large-scale data. Moreover, they frequently overlook the critical role of spatial context. VeloAgent addresses these issues by using agent-based simulations. This strategy ensures that the model considers how a cell's physical location influences its gene expression and future state. Consequently, the framework produces more biologically relevant results than traditional models that ignore spatial arrangement.

Predicting Disease Progression with Spatially-informed RNA Velocity

One of the most powerful features of veloAgent is its in silico perturbation module. Specifically, this module allows scientists to manipulate velocity vectors within a digital model. Researchers can therefore predict how targeted interventions might alter cell fate dynamics. This capability is vital for oncology, as it helps map the complex interactions within the tumor microenvironment. Furthermore, it offers a scalable way to test therapeutic hypotheses before moving to expensive laboratory experiments.

In addition to its predictive power, veloAgent achieves sublinear memory scaling. This efficiency enables the analysis of massive, multi-batch spatial datasets that were previously too large to process. Because it manages resources effectively, it serves as a versatile solution for researchers across diverse biological fields. By combining deep generative modeling with spatial awareness, veloAgent provides a robust roadmap for steering cellular dynamics toward desired outcomes.

Frequently Asked Questions

How does veloAgent improve RNA velocity accuracy?

VeloAgent integrates local microenvironment cues through agent-based simulations. This allows the model to account for spatial constraints that influence gene expression, leading to more precise inferences of cell state transitions.

What is the benefit of the in silico perturbation module?

This module enables researchers to digitally manipulate spatial velocity vectors. By doing so, they can simulate how specific regulatory changes or drugs might impact the direction of cell development and final cell fate.

Is veloAgent suitable for very large datasets?

Yes. The framework uses sublinear memory scaling, which allows it to process large-scale, multi-batch spatial datasets efficiently without requiring excessive computational resources.

Disclaimer: This content is for informational and educational purposes only. It is not a substitute for professional medical advice, diagnosis, or treatment. Always seek the advice of your physician or other qualified health provider with any questions you may have regarding a medical condition. Refer to the latest local and national guidelines for clinical practice.

References

Raghavan V et al. Dissecting and steering cell dynamics using spatially-informed RNA velocity with veloAgent. Mol Syst Biol. 2026 May 06. doi: 10.1038/s44320-026-00213-w. PMID: 42092184.

Yoon B et al. spVelo: RNA velocity inference for multi-batch spatial transcriptomics data. bioRxiv. 2025 Mar 11. doi: 10.1101/2025.03.11.641523.