Benchmarking the flow of epithelial cell monolayer with self-aligning deformable active membranes

arXiv:2608.13720 · cond-mat.soft · Submitted 2026-08-13 · Read on arXiv

Listen

Radio episode about this paper

Transcript

Introduction to the show: ident: Condensed Matter. The day in condensed matter: materials, phases and the measurements under them.

Kai: I'm Kai.

Mira: And I'm Mira, here with Lev.

Kai: We're looking at the paper: "Benchmarking the flow of epithelial cell monolayer with self-aligning deformable active membranes".

Mira: Here's the gist of the paper. Collective cell migration emerges from the interplay between motility, deformability and mechanical interactions <ref:2608.13720#pg6>.

Kai: This work benchmarks self-aligning deformable active membranes in a confined-flow geometry that mimics epithelial monolayer migration around a circular obstacle <ref:2608.13720#pg2>.

Mira: The model captures a broad spectrum of tissue behaviors, ranging from disordered, liquid-like flows to highly aligned, solid-like states <ref:2608.13720#pg6>.

Kai: Model Formulation The model employed is based on the active membrane model introduced by Teixeira et al.

Paper summary: Kai: So, we're looking at this paper here, "Benchmarking the flow of epithelial cell monolayer with self-aligning deformable active membranes". It’s about how cells move when they’re packed together and interacting mechanically.

Mira: Yeah, the main idea is that collective cell migration comes from a mix of how motile the cells are, how squishy they are, and how they push against each other physically. This work sets up a test case by looking at self-aligning deformable active membranes in a confined flow setting that mimics how epithelial sheets move around something like a circular obstacle.

Lev: It seems to capture a wide range of behaviors, from things that look like messy liquid flows to states that are really solid and highly aligned, which is interesting for understanding tissue structure.

Kai: Exactly. The authors use this model to test out how different physical settings affect the collective movement of these simulated cells around an obstacle. They’re checking if they can get a good description of real tissue behavior using this setup.

Mira: What’s interesting about their approach is that they are systematically testing three main things: velocity alignment, adhesion between membranes, and density. By setting these parameters to high or low values, they explore eight different scenarios to see how those factors play out.

Lev: I wonder how much control you actually have over those physical quantities in a real system. That sounds like it might be a lot of tuning for any experimental setup we try to build.

Kai: Right, that’s the kind of thing we’re thinking about when we design our own hardware. For instance, they use something called the target shape index p0 to control how deformable the membranes are and their shape.

Mira: And they found that increasing this target shape index p0 actually promoted cell elongation and made the tissue flow faster. This directly connects how deformable a single cell is to how fast the whole tissue moves, which is a key link.

Lev: So, if you make the cells more elongated in this model, the overall transport speed goes up. That means for real systems, we’d need to worry about that deformability scaling with flow rates.

Kai: It seems like they also managed to get their simulation results closer to both extremes of what we call the solid-liquid spectrum when compared against a related multiparticle model.

Mira: Compared to that other model, self-aligning active membranes actually achieved stronger collective alignment in their simulations. That’s a big finding because it suggests this specific mechanism is more effective at organizing the cells.

Lev: Stronger alignment means the tissue moves more coherently, which is important for any real biological function you want to replicate with hardware. But we need to be careful about how much of that alignment depends on those adhesion parameters they tuned.

Paper summary: Kai: Speaking of things that depend on tuning, the paper also compares their simulation results against experimental data along several lines around the obstacle. They found that the velocity profiles they simulated qualitatively matched what was actually seen experimentally in several directions.

Mira: But they did have some discrepancies, specifically when you look upstream of the obstacle and in how the transverse motion looked asymmetrical compared to what was measured experimentally.

Lev: That asymmetry is important because it tells us there’s a specific physical mechanism or boundary condition that their model isn't capturing perfectly right now. We need to figure out why that happens.

Kai: So, while the overall picture of collective flow looks plausible, the fine details—especially near boundaries and in specific motion directions—are where we need to focus our next efforts.

Mira: This paper is useful because it provides a clear benchmark for how these deformable membranes behave under different mechanical stresses. It gives us a way to visualize the spectrum of tissue behaviors they are modeling, from disordered flows to highly ordered states.

Lev: From an engineering standpoint, the fact that they can systematically vary those physical quantities and see the results map out what we need to measure in a real experiment is valuable for designing our testbeds. We know what kinds of measurements will give us the most information about deformation versus alignment.

Kai: The paper also suggests a way forward where we could introduce active regulation of membrane contractility by changing how tension is applied across the bonds spatially and temporally.

Mira: If you can modulate that tension, it opens up possibilities for understanding how cell deformability and active contractility work together to determine the final collective migration path. That moves us from just describing the flow to controlling it.

Lev: Controlling contractility through bond tension modulation sounds like a very difficult task for real hardware right now, but it’s a concept we have to work toward for any functional system. It’s about adding another layer of control on top of the existing alignment and density dynamics.

Kai: So, in short, this paper shows us that the interplay between motility, deformability, and mechanical forces is complex enough to generate a wide range of tissue behaviors, and it gives us a good framework to test those interactions.

Mira: The authors’ benchmarking work helps establish what we should expect when we look at epithelial migration in confined spaces. It defines the landscape of possibilities for this kind of modeling.

Lev: For anyone building something that tries to mimic this behavior, knowing where the model stops working or where the experimental data disagrees is just as important as knowing where it succeeds.

Kai: That’s what we’ll be looking at next, how we can take these insights about deformability and contractility and actually translate them into a system we can build and measure with high fidelity.

Conclusion: Kai: So this paper is setting up a benchmark for how we look at epithelial cells moving in groups around an obstacle using these self-aligning membranes.

Mira: Yeah, looking at the title, "Benchmarking the flow of epithelial cell monolayer with self-aligning deformable active membranes," it sounds like they’re trying to create a reliable test case for this whole area.

Lev: From where I'm sitting on error correction for real hardware, a benchmark is crucial because you need something you can actually build and measure against.

Kai: Exactly. They’ve built a model that lets them see the whole spectrum of tissue behavior, from liquid-like chaos to solid alignment states around that circular obstacle.

Mira: The authors are systematically testing three main physical knobs—velocity alignment, how much adhesion there is between the membranes, and the density of those cells.

Lev: I’m thinking about what this means for real systems; tuning those parameters isn't always straightforward when you try to build something that actually works in a lab.

Kai: Right, they found that controlling things like the target shape index which dictates cell deformability really boosted both the cell elongation and how fast the whole tissue flow went.

Mira: So, if you can make your cells stretch out more easily, it directly affects how quickly the tissue moves collectively around something physical. That links single-cell mechanics to tissue transport.

Lev: For us on error correction, that means we need to know exactly what kind of mechanical input is driving that alignment so we can design better control mechanisms for real hardware.

Kai: They did compare their simulation results to actual experimental profiles around the obstacle, and they got a qualitative match in several directions.

Mira: But there were still specific discrepancies, like the motion upstream of the obstacle and some asymmetry in how it moved sideways compared to what people actually saw experimentally.

Lev: That asymmetry is where you know your model is missing something—it’s a boundary condition or a physical interaction they haven't captured yet.

Kai: The paper points toward using active regulation of membrane contractility, which means changing the tension in the bonds based on space and time.

Mira: If they can modulate that tension, it opens up a way to study how cell deformability and active contractility work together to decide the path of collective migration.

Lev: It moves the problem from just describing flow to actually controlling it, which is a bigger hurdle for us when we try to implement these ideas on actual hardware.

Kai: So this whole piece is essentially mapping out the mechanical landscape of how cells move in tight groups, and showing us where the current models succeed and where we need better physics.

Mira: It gives us a solid framework for what to expect when we look at epithelial behavior under mechanical stress.

Lev: It also shows us that while the basic idea is sound, there’s still a lot of fine-tuning required before this becomes something you can run on a real quantum processor or experimental setup.

Marcos Pasaa, Carine P. Beatricia, François Granerc, Leonardo G. Brunneta, Emanuel F. Teixeirab

Instituto de Física, Universidade Federal do Rio Grande do Sul · Huygens-Kamerlingh Onnes Laboratory, Universiteit Leiden · Université Paris Cité, CNRS

cond-mat.soft

Submitted: 2026-08-13

Updated: 2026-11-02

Comments: 16 pages, 14 figures v3: Added a missing link to the supplementary videos. No changes to the results or conclusions

License: http://creativecommons.org/licenses/by/4.0/

Importance score: 83/100

The gist: Collective cell migration emerges from the interplay between motility, deformability and mechanical interactions <ref:2608.13720#pg6>.

Key concepts

Active Membrane Model (DAMM)
This is a simulation framework based on active membranes where each cell boundary is treated independently. It models cells as collections of particles connected by elastic bonds, allowing the system to exhibit self-propulsion and complex tissue behaviors driven by internal energy functions.
Deformability (Target Shape Index p0)
This parameter controls how easily the membrane can stretch and change its shape. Increasing this index promotes cell elongation, which directly links how cells deform with the overall movement of the tissue flow around an obstacle.
Solid–Liquid Spectrum
This refers to the range of behaviors observed in the simulation, spanning from disordered, liquid-like flows where membranes move randomly to highly aligned, solid-like states where they move coherently and systematically.

Terminology

Summary

Collective cell migration emerges from the interplay between motility, deformability and mechanical interactions <ref:2608. p. 0

Where the paper says it

Brunneta and Emanuel F. Teixeirab Collective cell migration emerges from the interplay between motility, deformability and mechanical interactions, yet incorporating these ingredients into computationally efficient tissue models remains challenging.

13720#pg6>. This work benchmarks self-aligning deformable active membranes in a confined-flow geometry that mimics epithelial monolayer migration around a circular obstacle <ref:2608. p. 0
Where the paper says it

Teixeirab Collective cell migration emerges from the interplay between motility, deformability and mechanical interactions, yet incorporating these ingredients into computationally efficient tissue models remains challenging. Here, we benchmark self-aligning deformable active membranes in a confined-flow geometry that mimics epithelial monolayer migration around a circular obstacle.

13720#pg2>. The model captures a broad spectrum of tissue behaviors, ranging from disordered, liquid-like flows to highly aligned, solid-like states <ref:2608. p. 0
Where the paper says it

By systematically varying the self-alignment timescale, cell–cell adhesion and inlet forcing, we characterize the resulting flows through collective alignment, relative density, neighbor rearrangements and spatial velocity fields. The model captures a broad spectrum of tissue behaviors, ranging from disordered, liquid-like flows to highly aligned, solid-like states.

13720#pg6>.

Model Formulation

The model employed is based on the active membrane model introduced by Teixeira et al. 47 <ref:2608. p. 0

Where the paper says it

To capture the geometry of confluent tissues, vertex models represent cells as polygonal domains whose vertices constitute the fundamental degrees of freedom25–27, whereas JournalName,[year],[vol. ],1–18 | 1 arXiv:2608.

13720#pg4>. The system consists of N active membranes, each composed of n active particles connected by elastic bonds <ref:2608. p. 2
Where the paper says it

2 Model The model employed in this work is based on the active membrane model introduced by Teixeira et al. The system consists of N active membranes, each composed of n active particles connected by elastic bonds.

13720#pg4>. A distinctive feature of the DAMM is its explicit representation of cell boundaries, assigning an independent membrane to each cell <ref:2608. p. 1
Where the paper says it

A distinctive feature of the DAMM is its explicit representation of cell boundaries. Unlike Vertex and Voronoi-based models, where neighboring cells share a single geometric interface, the DAMM assigns an independent membrane to each cell, allowing interactions to occur directly between adjacent cell membranes.

13720#pg5>. The global polarity determines the overall self-propulsion direction of the membrane and aligns with its center-of-mass velocity rather than with the velocity of each individual particle <ref:2608. p. 2
Where the paper says it

Unlike earlier implementations in which each constituent particle possessed an independent polarity 39,46, all particles within a given membrane share a common polarity, as in Refs. This global polarity determines the overall self-propulsion direction of the membrane and aligns with its center-of-mass velocity rather than with the velocity of each individual particle.

13720#pg5>. The energy function has contributions from a perimeter energy (EP) (springs connecting membrane neighboring particles), preferred-area energy (EA), core repulsion, and intercellular adhesion among particles (Ec,adh) <ref:2608. p. 2
Where the paper says it

The noise term DR represents the rotational diffusion coefficient, which defines a characteristic timescale given by τR = 1/DR. The energy function has contributions from a perimeter energy (EP) (springs connecting membrane neighboring particles), preferred-area energy (EA), core repulsion, and intercellular adhesion among particles (Ec,adh): EP = kP 2 N ∑ j=1 n ∑ i=1 li, j −l0 2 , (2a) EA = kA 2 N ∑ j=1 Aj −A0 2 , (2b) Ec,adh = kc 2 ∑ rik≤σ (rik −σ) 2 + kadh 2 ∑ σ<rik≤ladh (rik −σ) 2 , (2c) E = EP +EA +Ec,adh.

13720#pg5>.

Parameter Exploration Strategy

The authors systematically varied three physical quantities of interest—Velocity alignment (A), Adhesion between membranes (T), and Density (D)—by setting model parameters to either high or low values <ref:2608.13720#pg6>. The parameter ranges were constrained by numerical limitations and physical considerations, leading to 2 cubed = 8 possible scenarios to simulate <ref:2608. p. 4

Where the paper says it

We ran simulations for all combinations of the selected parameter extremes. The parameter ranges are constrained by numerical limitations and physical considerations.

13720#pg6>. For instance, the target shape index p0 controls membrane deformability and shape <ref:2608. p. 1
Where the paper says it

We show that, under adequate parametrization, the DAMM promotes strong spatial cohesion and robust collective ordering. By systematically varying the target shape index p0, which controls membrane deformability and shape, we demonstrate that increased deformability leads to faster flow, highlighting the central role of mechanical properties in collective migration.

13720#pg5>.

Key Findings from Simulations

The DAMM achieved higher alignment and exhibited a more systematic density response compared to a related multiparticle model <ref:2608.13720#pg6>. Increasing the target shape index p0 promoted cell elongation and accelerated tissue flow, directly linking cell-scale deformability to tissue-scale transport <ref:2608. p. 0

Where the paper says it

Compared with a related multiparticle model, self-aligning active membranes achieve stronger collective alignment, exhibit a more systematic density response and access states closer to both limits of the solid–liquid spectrum. We further show that increasing the target shape index promotes cell elongation and accelerates tissue flow, directly linking cell-scale deformability to tissue-scale transport.

13720#pg6>. The DAMM reached values closer to both limits of the solid–liquid spectrum compared to the MPM <ref:2608. p. 9
Where the paper says it

Rapid self-alignment is the main mechanism promoting collective order, while adhesion and inlet forcing modulated tissue cohesion, density, and neighbor rearrangements. Compared with the Multiparticle Model (MPM), the DAMM reached substantially higher alignment, exhibited a more systematic increase in density with inlet forcing, and accessed states closer to both limits of the solid–liquid spectrum.

13720#pg6>.

Comparison with Experiment

The simulated velocity profiles qualitatively reproduced experimental profiles along several horizontal and vertical transects around the obstacle <ref:2608.13720#pg6>. The principal discrepancies occurred upstream of the obstacle and in the asymmetric transverse motion observed experimentally <ref:2608.13720#pg6>.

Future Directions

The DAMM provides a natural framework where active regulation of membrane contractility could be introduced through spatial and temporal modulation of the tension of membrane bonds <ref:2608.13720#pg6>. Incorporating these feedbacks would allow the DAMM to address how cell deformability, active contractility, chemical signaling, and environmental mechanics jointly determine collective migration <ref:2608. p. 0

Where the paper says it

Teixeirab Collective cell migration emerges from the interplay between motility, deformability and mechanical interactions, yet incorporating these ingredients into computationally efficient tissue models remains challenging. Here, we benchmark self-aligning deformable active membranes in a confined-flow geometry that mimics epithelial monolayer migration around a circular obstacle.

13720#pg6>.

Green — the paper says it, and the page is printed. Red — no passage carries it, so treat that sentence as unverified. The same check runs on your own text: check a summary against its source, $5.

Improvements for AI systems

  1. Model enhanced simulation capabilities by incorporating explicit cell boundaries using Deformable Active Membrane Model (DAMM). This allows AI systems to capture cell-shape adaptations that are essential in biological tissues by modeling how increased deformability leads to faster flow.

  2. Develop predictive models for tissue transport velocity based on cell shape parameters. The improved system can calculate the mean flow speed v up to 10 times v0, showing that more deformable cells can accommodate local stresses and modify their shapes while moving through crowded regions, thereby facilitating rearrangements and collective transport.

  3. Create robust benchmarks for active matter models against experimental data. The AI system can now validate its predictions by comparing simulated velocity profiles with experimental measurements from in vitro migrating MDCK epithelial cell monolayers across multiple transects around an obstacle.

  4. Implement a mechanism to assess tissue organization states using solid-liquid behavior metrics. The system can determine if the tissue is exhibiting liquid or solid behavior by calculating the contribution to delta as defined in equation (10), accessing states closer to both limits of the solid–liquid spectrum.

  5. Improve flow prediction accuracy by tuning parameters that control alignment and adhesion. The AI can systematically vary target shape index p0, noting that increasing the target shape index promotes cell elongation and accelerates tissue flow, linking cell-scale deformability directly to tissue-scale transport.

  6. Enhance spatial field characterization using coarse-graining techniques for input and output measurements. The system can generate smooth velocity fields by applying a Gaussian filter and interpolation, allowing for more reliable analysis of Relative Density Field (delta rho) and Velocity Field (v).

  7. Establish uncertainty quantification for simulation results by exploring parameter sensitivity. By running simulations systematically varying tau, Fi, and kadh around a reference point, the system can provide an estimate of the sensitivity of the input measurements to each varied model parameter.

Abstract

Collective cell migration emerges from the interplay between motility, deformability and mechanical interactions, yet incorporating these ingredients into computationally efficient tissue models remains challenging. Here, we benchmark self-aligning deformable active membranes in a confined-flow geometry that mimics epithelial monolayer migration around a circular obstacle. In this model, cells are represented as deformable, adhesive membranes whose self-propulsion direction relaxes towards their velocity. By systematically varying the self-alignment timescale, cell-cell adhesion and inlet forcing, we characterize the resulting flows through collective alignment, relative density, neighbor rearrangements and spatial velocity fields. The model captures a broad spectrum of tissue behaviors, ranging from disordered, liquid-like flows to highly aligned, solid-like states. Compared with a related multiparticle model, self-aligning active membranes achieve stronger collective alignment, exhibit a more systematic density response and access states closer to both limits of the solid-liquid spectrum. We further show that increasing the target shape index promotes cell elongation and accelerates tissue flow, directly linking cell-scale deformability to tissue-scale transport. Finally, we compare simulated velocity profiles with experimental measurements from in vitro migrating MDCK epithelial cell monolayers and find qualitative agreement across multiple horizontal and vertical transects around the obstacle. These results establish self-aligning deformable active membranes as a versatile framework for connecting cell mechanics, shape adaptation and self-alignment to collective tissue migration in confined geometries.

Sources

Related papers