Germinal centers are dynamic immune structures in which B cells repeatedly alternate between a proliferative dark zone and a selection-dominated light zone. During this cycle, centroblasts divide and diversify their immunoglobulin receptors, while selected centrocytes can receive survival signals, return to the dark zone and undergo further rounds of mutation and selection. Some cells eventually leave the reaction as prememory B cells or plasmablasts, which can give rise to memory B cells or plasma cells. Because single-cell sequencing destroys the cells it profiles, the sequence and regulation of these transitions must be inferred from a static snapshot, making it difficult to distinguish cell-fate drivers from correlations and to predict how a mutation will alter trajectories.

DynaVelo addresses that problem by treating RNA expression and transcription-factor motif accessibility as two connected dynamical systems. Variational encoders first map the two data types into lower-dimensional latent spaces, where a shared neural ordinary differential equation learns a vector field describing cellular change over an inferred latent time. Separate decoders translate that field back into gene-expression and motif-accessibility space, allowing the model to estimate RNA velocities for all modeled genes as well as velocities for transcription-factor motifs. Training combines reconstruction of the observed data, regularization of initial states and latent times, alignment with available scVelo RNA-velocity estimates for a subset of genes, and a consistency term requiring RNA and motif velocities to imply similar trajectories. The resulting model is generative, so it can create cells along inferred trajectories and simulate loss-of-function changes without requiring new cells for every perturbation.

Applied to 7,166 wild-type mouse germinal-center B cells, DynaVelo reconstructed the major branches of the reaction. Its model-derived vector field captured the centroblast cell cycle, the transition from centroblasts to light-zone centrocytes, the return of selected centrocytes to the dark zone, and exits toward prememory cells and plasmablasts. Its latent-time ordering more closely followed these biological transitions than the ordering produced by scVelo, which did not identify the bifurcation between dark-zone cycling and light-zone differentiation. A comparison with the multiome velocity method MultiVelo likewise found differences: MultiVelo did not recover the dominant centroblast-to-centrocyte direction and placed plasmablasts near the beginning rather than at a terminal exit. These results describe dynamics inferred from sampled cells and model outputs, not direct continuous observation of cells moving between zones.

Generated trajectories also exposed cell-state-specific regulatory patterns. Motif accessibility for Spi1 and Nfkb1 increased during the centroblast-to-centrocyte transition, whereas Spi1 remained stable and Nfkb1 decreased during dark-zone cycling. Along the route toward prememory cells, the model identified temporal changes in factors including Runx1 and Irf8, while the plasmablast branch showed accessibility and expression patterns involving Prdm1, Irf4 and Xbp1. The distinction between expression and motif accessibility was important: Prdm1 motifs were accessible in both prememory and plasmablast states, but Prdm1 RNA was expressed only in plasmablasts. The analysis therefore separated RNA-expression patterns from the accessibility of regulatory motifs, while the factors identified as possible drivers remain interpretations of temporal patterns rather than proof of causation in every inferred transition.

The same framework revealed how heterozygous deletions of Arid1a or Ctcf altered the inferred germinal-center landscape in samples containing 6,716 and 4,688 cells, respectively. Both mutants retained recognizable cell-cycle and differentiation branches but lacked the light-zone-to-dark-zone recycling path. Prememory cells accounted for 5.42% of wild-type cells, compared with 20.09% in Arid1aHet and 25.45% in CtcfHet germinal centers; the study also examined an Arid1a homozygous sample, in which the proportion was 44.96%. These are differences among genotypes, not measurements presented as a decline over time. The model associated the loss of recycling with altered activity of factors including Foxp1, Bach2, Spi1 and Ctcf, and with loss of a Foxo1-to-Cxcr4 regulatory relationship associated with migration back to the dark zone.

DynaVelo inferred dynamic gene-regulatory networks through two related calculations. Jacobian matrices measured the local derivative of one gene’s RNA velocity with respect to another gene’s expression, revealing how regulatory relationships varied among individual cells and along trajectories. Finite in silico knockouts instead set a transcription factor’s RNA and motif accessibility to their minimum observed values and calculated the resulting changes in velocity. The two approaches were not identical: Jacobians represented infinitesimal local changes, whereas simulated knockouts imposed finite changes across cells. Together, they recovered multiple previously reported relationships in wild-type cells while identifying edges that were missing or changed in the mutants. One example was the Spi1-to-Bcl-6 relationship, which disappeared along the centroblast-to-centrocyte-to-prememory trajectory in both heterozygotes. The sign of the modeled Bcl-6 influence on Bcl-2 also differed between wild-type and mutant cells.

To evaluate the model as a predictor of genetic loss, the researchers simulated Arid1a loss in wild-type cells and compared the predicted velocity changes with differential velocities inferred from Arid1aHet cells. Agreement was generally stronger for motif velocities than for RNA velocities across cell types. The simulation predicted reduced connectivity between centroblasts and centrocytes, loss of recycling, and a stronger route from centrocytes to prememory cells, matching the altered mutant structure. The study then perturbed candidate genes in the Arid1aHet model and searched for changes in RNA and motif velocities that pointed toward wild-type values. Ctcf, Stat3 and Bcl6 were among the strongest candidates, alongside Tox and Bcor. A velocity was counted as rescued when its direction of change matched the direction from mutant to wild-type in that cell state, making these computational predictions distinct from the subsequent in vivo outcomes.

Experimental profiling of Arid1aHet;CtcfHet double-heterozygous germinal centers showed a partial reversal of the rapid centrocyte-to-prememory exit characteristic of the Arid1a mutant. In centrocytes, the measured rescue rates were 38% for RNA expression, 64% for RNA velocity, 51% for motif accessibility and 46% for motif velocity. The defining Arid1aHet phenotype was an elevated dark-zone/light-zone ratio. Two newly generated compound-heterozygous models, Arid1aHet;Bcl6Het and Arid1aHet;Stat3Het, each restored that ratio to wild-type levels, providing direct functional support for two predictions. Together with the Ctcf double heterozygote, the experiments validated all three nominated targets. The results show that DynaVelo can nominate experimentally testable genetic interactions in this mouse system, but the study does not establish clinical effectiveness, and its predictions remain dependent on the quality of the training data, chosen genes, velocity guidance and model settings.

The experimental work used age- and sex-matched mice, with both males and females represented in each group, and the researchers reported no sex-dependent effects. Germinal-center B cells were profiled by paired single-cell RNA sequencing and ATAC sequencing, while flow cytometry measured cell-surface markers and the dark-zone/light-zone composition of germinal centers. The computational inputs included 2,000 highly variable genes, approximately 400–1,000 genes with scVelo velocity estimates, about 200 expressed transcription factors and roughly 20 genes of interest, with the union used for modeling. The study used separate wild-type, Arid1aHet, CtcfHet and compound-heterozygous samples, and it trained each model on a 90:10 training and validation split with early stopping. These design features support the model’s use as a trajectory and perturbation-analysis framework for this particular germinal-center system, while the experimental rescue results concern mouse phenotypes rather than human disease outcomes or therapeutic interventions.

Subject of Research: Computational modeling of mouse germinal-center B-cell dynamics and validation of predicted genetic rescue targets

Article Title: Deep dynamical models of single-cell multiomic velocities predict loss-of-function and rescue perturbations in B cells

Article References: Karbalayghareh, A., Pelzer, B., Chin, C. R., Melnick, A., Barisic, D., & Leslie, C. S. (2026). Deep dynamical models of single-cell multiomic velocities predict loss-of-function and rescue perturbations in B cells. Nature Biotechnology. https://doi.org/10.1038/s41587-026-03348-1

Image Credits: AI Generated

DOI: 10.1038/s41587-026-03348-1

Keywords: single-cell multiome, germinal-center B cells, neural differential equations, gene-regulatory networks, in silico perturbation, Arid1a, Ctcf, Bcl6, Stat3, epigenetic regulation, Deep, dynamical

The crux of biological systems lies in the balance between deterministic rules—governing cellular processes in predictable ways—and stochastic events, which introduce variability and randomness that can significantly impact cell fate. Traditional models have primarily focused on deterministic dynamics while treating diffusion, or the random fluctuations, as constant. However, through scDiffEq, researchers can now develop a nuanced understanding of how these factors contribute to the diversity of cellular outcomes observed during crucial stages such as development and responses to therapies.

To demonstrate the utility of scDiffEq, the researchers employed lineage-traced single-cell data, offering a unique perspective on how multipotent progenitors navigate their fates during haematopoiesis—the formation of blood cells. Their findings illustrated a dramatic improvement in reconstructing cell trajectories and predicting the ultimate fate of these cells as they differentiate into specialized types. The implications of such advancements could lead to more precise therapeutic interventions in various diseases, particularly those involving hematologic disorders.

The study goes on to explore the fascinating capabilities of scDiffEq by introducing in silico perturbations to the multipotent progenitor cells. Through simulations mimicking CRISPR-based genome editing, the researchers were able to replicate the dynamic responses seen in real experimental conditions. These findings provide critical validation for the scDiffEq model, suggesting that it not only accurately reflects biological processes but also retains the flexibility and robustness necessary for application across different contexts.

Additionally, the researchers break new ground by extending the scope of scDiffEq beyond lineage-tracing and multi-time-point datasets. This adaptability allows for the modeling of single-cell data obtained at only one point in time. The significance of this capability cannot be overstated, as it opens avenues for analyzing a wide array of datasets that were previously difficult to interpret. In particular, scDiffEq is positioned to advance the study of cellular dynamics in systems where time-course data may be scarce or unavailable.

Central to the power of scDiffEq is its ability to recreate high-resolution developmental cell trajectories. By accurately modeling both the drift—the predictable nature of cellular development—and the underlying diffusion—the stochasticity—researchers can gain deeper insights into time-dependent gene-level dynamics. This has enormous potential for a variety of applications, from understanding the cellular intricacies of early development to dissecting the cellular mechanisms involved in cancer progression.

The advantages of scDiffEq are not limited to modeling techniques. By skillfully integrating advanced computational methods with biological modeling, the research exemplifies how interdisciplinary approaches can profoundly enhance scientific discovery. The use of neural networks to approximate the underlying dynamics of cellular behavior represents a novel fusion of biology and artificial intelligence. This could herald a new era in predictive modeling, providing scientists with tools to envision and manipulate cellular behaviors through simulations.

Furthermore, the researchers shed light on the implications of these findings for potential therapeutic strategies. With scDiffEq poised to bridge the gap between understanding basic biological principles and applied biotechnology, it could enable more effective strategies for manipulating cell fate in regenerative medicine and cancer therapeutics. By accurately predicting how cells will respond to various perturbations, scDiffEq could facilitate the design of interventions that promote desirable outcomes in tissue regeneration or disease correction.

As this research illustrates, the journey toward decoding the complexities of cellular dynamics is only just beginning. The partnership between experimental biology and computational modeling stands to redefine the landscape of our understanding. With scDiffEq, the research community gains not only a powerful tool but also a framework that fosters collaboration, encourages exploration, and ultimately drives the evolution of cell biology.

The implications of this research extend beyond immediate applications; they also resonate with broader themes in science. As researchers embark on this deeper exploration of cellular dynamics, there is an overarching emphasis on the need for tools and frameworks that can adapt to the rapidly evolving landscape of biological data. In this context, scDiffEq is not merely an improvement upon existing models; it represents a paradigm shift in how we conceptualize biological systems.

In conclusion, the advent of scDiffEq signifies a pivotal moment in the intersection of biology, technology, and data science. By addressing historical limitations in modeling cellular dynamics and providing a robust platform for future research, this study sets the stage for groundbreaking advancements in understanding the intricate tapestry of life at the cellular level. As scientists harness the full potential of this innovative framework, we can anticipate a new wave of discoveries that will unravel the mysteries of development and disease, paving the way for a future enriched with improved diagnostic and therapeutic tools.

Researchers are encouraged to adopt and adapt scDiffEq in their own studies, fostering a broader dialogue on the role of stochastic dynamics in biology. The future is bright as we stand at the precipice of potentially transformative advancements in our understanding of life itself.

Subject of Research: Cellular dynamics and stochastic differential equations

Article Title: Learning cell dynamics with neural differential equations

Article References:

Vinyard, M.E., Rasmussen, A.W., Li, R. et al. Learning cell dynamics with neural differential equations.
Nat Mach Intell (2025). https://doi.org/10.1038/s42256-025-01150-3

Image Credits: AI Generated

DOI: https://doi.org/10.1038/s42256-025-01150-3

Keywords: cell dynamics, stochastic differential equations, computational biology, haematopoiesis, CRISPR, artificial intelligence, predictive modeling