On a Keller-Segel type equation to model Brain Microvascular Endothelial Cells growth's patterns

arXiv:2604.25180 · math.DS, q-bio.QM · Submitted 2026-04-28 · Read on arXiv

Listen

Transcript

Introduction to the show: ident: Genomics Radio. Generated commentary on the latest computational biology and genomics papers.

Ines: Today's paper: "On a Keller-Segel type equation to model Brain Microvascular Endothelial Cells growth's patterns".

Marcus: The gist The article presents a partial differential equation (PDE) of Keller–Segel (KS) type that reproduces patterns commonly observed during the growth of brain microvasculature,

Ines: First, who's behind it and why it matters.

Paper summary: Ines: So, to wrap up this paper, they introduced this modified KS model that successfully reproduces patterns seen in in vitro BMEC growth experiments and gives us mathematical insights into the mechanisms driving that pattern formation.

Marcus: They also computed distributions of chemoattractant concentrations that are consistent with the time evolution of those experimental growth patterns based on their model. It’s a big piece of work because it connects the math to what we see in culture.

Yuki: The authors are doing this by proposing this KS-type equation and then showing how variations in parameters lead to network-like spatial structures that match what was seen in experiments.

Ines: They also propose the data-driven extension for v using consecutive images to ensure a consistent temporal evolution of the chemoattractant. That's how they bridge the gap between static patterns and dynamic biological processes.

Marcus: The whole point is building this comprehensive mathematical modeling framework spanning blood flow, biochemical processes, and cell behavior to better understand vascular impairments contributing to neurodegenerative diseases.

Yuki: It sets up some in silico platforms that could eventually support the development of new therapeutics for neurodegenerative diseases by identifying the molecular drivers of those pathological interactions.

Conclusion: Ines: So, we've looked at this modified Keller–Segel equation and seen how it actually spits out patterns that look like brain blood vessels growing in culture.

Marcus: Yeah, the authors are basically saying they have a mathematical tool now to figure out *why* those specific shapes form when endothelial cells are growing.

Yuki: From my angle, the real question is whether these observed patterns are just random noise or if there's a genetic blueprint in how these cells interact chemically.

Ines: Exactly. They’re using this equation to show that changes in one parameter, like the growth rate gamma, directly control whether you get a network of structures or something else entirely.

Marcus: And they even built some data-driven stuff, treating two microscopy images as input to figure out how the chemoattractant field v actually evolves over time.

Yuki: That's the part that really connects to the real world—it’s not just a neat equation on paper; it’s trying to reconstruct the actual biological timing of those cell movements.

Ines: The implication here is they are building a bigger framework, trying to connect how blood flows in arterial networks with all these biochemical signals.

Marcus: It moves us toward using these computational models as tools to predict how vascular problems might feed into neurodegenerative diseases down the road.

Yuki: So we're looking at using this math to create these in silico platforms, right? Systems that could help us test new treatments for conditions like Alzheimer's or Parkinson's.

University Le Havre Normandie · The Hudson School of Mathematics · University Sorbonne Paris Nord · School of Medicine, Conway Institute, University College of Dublin, Ronald M Loeb Center for Alzheimer’s Disease, Friedman Brain Institute, Glickenhaus Center for Successful Aging, Icahn School of Medicine at Mount Sinai

math.DS, q-bio.QM

Submitted: 2026-04-28

Updated: 2026-10-14

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

Importance score: 76/100

The gist: The gist The article presents a partial differential equation (PDE) of Keller–Segel (KS) type that reproduces patterns commonly observed during the growth of brain microvasculature, providing

Key concepts

Keller-Segel (KS) Type Equation: This is a mathematical model used to describe how cells move and aggregate based on chemical signals. In this paper, it's modified to specifically mimic the growth patterns observed in brain microvascular endothelial cells (BMECs). It helps explain the mechanisms driving how these cells organize themselves spatially.
Chemoattractant Concentration ($v$): This variable represents the concentration of a chemical signal that attracts endothelial cells. The model uses this to show how the presence and movement of these cells are linked to the surrounding biochemical environment. Reconstructing its temporal evolution is key to matching experimental growth patterns.
Pattern Formation ($\gamma$ parameter): The parameter $\gamma$ controls a key aspect of cell behavior, influencing whether the system forms specific spatial structures. Varying this parameter between 0 and 1 allows researchers to observe how different values lead to the creation of network-like patterns resembling those seen in BMEC growth experiments.

Summary

The gist The article presents a partial differential equation (PDE) of Keller–Segel (KS) type that reproduces patterns commonly observed during the growth of brain microvasculature, providing mathematical insights into the mechanisms underlying these patterns.

Model Formulation

The study introduces a modified KS type equation to model endothelial cell growth, which is given by Equation (2): ut = f(u) − b∇ · (u∇v) + du∆u and vt = cu − ev + dv∆v The variable u represents the concentration of endothelial cells and v represents the substrate and the biochemical mechanisms through which cells communicate The reaction function is defined as f(u) = au(1 − u)(u − γ), where f(u) is relevant to model patterns formed by BMEC This equation is of Keller-Segel type with a modified reaction function

Mathematical Analysis

The basic mathematical analysis of the underlying ODE (Equation 4) reveals three stationary points: U∗0 = (0, 0), U∗1 = (γ, c e γ), and U∗2 = (1, c e) Proposition 1 states that U∗0 and U∗2 are stable nodes, while U∗1 is a saddle-node Furthermore, the solutions of Equation (2) remain non-negative, which is verified by Proposition 2 The average concentration values satisfy specific equations, as stated in Proposition 3.

Numerical Simulations and Pattern Emergence

Numerical simulations of the modified KS model (Equation 2) illustrate how variations in key parameters lead to the emergence of network-like spatial structures resembling experimentally observed BMEC patterns The parameter γ is varied between 0 and 1, and choosing γ in the range 0.25–0.29 leads to the formation of angiogenic patterns For instance, simulations for γ = 0.29 show that the solution initially evolves rapidly toward a state with both blue and black regions, leading to a structure similar to those observed in experiments.

Data-Driven Reconstruction

A data-driven extension of the model is proposed to ensure a consistent temporal evolution of the chemo-attractant associated with the observed BMEC dynamics This involves treating two consecutive microscopy images, u(t) and u(t + 1), as given data and computing the corresponding field v by solving Equation (9). This elliptic partial differential equation for v is crucial for reconstructing the temporal evolution of the cell patterns

Low-Dimensional Prototype

A two×two coupled ODE model (Equation 10) was proposed to investigate mathematical mechanisms, mimicking Equation (3) by incorporating diffusive terms and growth factors This low-dimensional system admits stationary solutions that represent two high concentrations of cells, no cells, and coexistence of high and low concentration of cells The analysis shows that stationary solutions U∗8 and U∗9 provide illustrations for the effects of nonzero growth factors, diffusive factors and chemotactic factors at equilibrium This system recovers the principles that led to the formation of BMEC patterns in the KS PDE.

Conclusion

In this article, a modified KS model is introduced that reproduces patterns observed in in vitro BMEC growth experiments and provides mathematical insights into the mechanisms driving this pattern formation Based on this model, distributions of chemoattractant concentrations consistent with experimental growth patterns have been computed. This research is part of a broader effort that includes developing biochemical mathematical models for angiogenesis, in vitro models incorporating BMECs, astrocytes, and neurons, as well as three-dimensional models of blood flow. The work aims to create in silico platforms to support the development of new therapeutics for neurodegenerative diseases.

How it works

The study employs a three-time-phase structure for γ = 0.25, distinguishing between very fast, fast and slow time scales At a fixed spatial position where u is asymptotically close to its maximum value, the dynamics exhibit three phases: an initial rapid increase where ∆u is positive and contributes to the fast increase in u, followed by a second phase where u increases slower as ∆u becomes negative, and a third phase where all quantities reach a stationary state, with negative contributions from growth and diffusive terms compensated by the chemotaxis term.

Numerical Implementation

To solve the elliptic equation for the field v, the Generalized Minimal Residual (GMRES) method is adapted because discretizing the operator leads to a sparse linear system that is numerically challenging GMRES constructs an approximate solution by minimizing the residual norm at each iteration. The algorithm involves computing matrix rotations Gk and solving an upper triangular system to compute the approximation of v, Xk = X0 + Vy.

Parameter Dependence

The dynamical behavior in Equation (10) is highly dependent on the parameter b, where as b increases, new unstable stationary points appear Specifically, for b = 25, four more stationary solutions U6∗ and U7∗ appear alongside two locally asymptotically stable solutions U8 and U9. The stability of these points is determined by the interplay between the growth factor, diffusive factors, and chemotactic factors at equilibrium.

Biological Relevance

The model aims to better understand how vascular impairments may contribute to neurodegenerative diseases by spanning blood flow in cerebral arterial networks to biochemical processes The use of induced pluripotent stem cells (iPSCs) differentiated into brain microvascular endothelial cells (BMECs) provides a framework for identifying the molecular drivers of pathological cell–cell interactions within this system. This mathematical modeling framework is intended to support the development of new therapeutics for neurodegenerative diseases.

The paper successfully introduces a modified KS model that reproduces patterns observed in in vitro BMEC growth experiments and provides mathematical insights into the mechanisms driving this pattern formation. Based on this model, distributions of chemoattractant concentrations consistent with the time evolution of the experimental growth patterns have been computed. This research is part of a broader effort that includes developing biochemical mathematical models for angiogenesis, in vitro models incorporating BMECs, astrocytes, and neurons, as well as three-dimensional models of blood flow. The work aims to create in silico platforms to support the development of new therapeutics for neurodegenerative diseases.

Numerical Implementation

Improvements for AI systems

Bold header: Improved Pattern Recognition in Neurovascular Health

This system can analyze microscopy images of Brain Microvascular Endothelial Cells (BMECs) and predict their spatial patterns by fitting them to the solutions of Equation (2), specifically identifying configurations corresponding to different values of the parameter γ, such as a network that suggests a stable, minimal-energy configuration or a degenerate network of cells.

Bold header: Real-Time Chemoattractant Dynamics Estimation

The AI can reconstruct the temporal evolution of the chemoattractant field 'v' using Equation (9), which allows it to derive the field 'v' from two consecutive microscopy images, thereby modeling the temporal evolution of cellular patterns based on experimentally observed data.

Bold header: Multi-Scale Dynamical Behavior Analysis

The system can distinguish between very fast, fast and slow time scales in the dynamics by analyzing the time evolution of quantities like 'u', '∆u', and '∇u · ∇v' at fixed spatial points, allowing it to characterize the three phases observed during pattern evolution.

Bold header: Low-Dimensional Prototype Simulation

The system can simulate the behavior of a simplified two×two coupled ODE model (Equation 10) under varying parameters 'b', determining when the system admits four new stationary points and identifying which solutions are locally asymptotically stable, such as those representing coexistence of high and low concentration of cells.

Bold header: Parameter Sensitivity Mapping

The AI can map how variations in key parameters, like 'γ' or 'b', affect the resulting spatial structures, quantifying the transition from a spatially homogeneous steady state to complex patterns like those observed for γ = 0.29, which shows a mixture of blue and black regions.

Bold header: Numerical Solver Robustness

The system can employ an adapted GMRES (Generalized Minimal Residual) method to solve the elliptic PDE for 'v', ensuring the numerical stability necessary for computing fields in high-resolution spatial discretizations where the original problem numerically remains challenging.

Abstract

This article presents a partial differential equation (PDE) of Keller-Segel (KS) type that reproduces patterns commonly observed during the growth of brain microvasculature. We provide mathematical insights into the mechanisms underlying the emergence of these patterns. In addition, we derive a data-driven equation that ensures a consistent temporal evolution of the chemoattractant associated with the observed microvascular dynamics. Beyond numerical simulations, the aim of this study is to advance a comprehensive mathematical modeling framework, spanning blood flow in cerebral arterial networks to biochemical processes, in order to better understand how vascular impairments may contribute to neurodegenerative diseases.