Towards Learned Simulators for Cell Migration

TL;DR

This paper introduces a neural probabilistic model for simulating single cell migration, offering faster and more accurate stochastic dynamics compared to traditional Cellular Potts models, with improved training strategies for stability.

Contribution

It presents a novel autoregressive neural simulator for cell migration that addresses training stability and stochastic accuracy issues, outperforming existing models in speed and fidelity.

Findings

Neural simulator reproduces cell migration dynamics accurately.

Training strategies improve stability and stochastic fidelity.

Simulation is at least ten times faster than Cellular Potts model.

Abstract

Simulators driven by deep learning are gaining popularity as a tool for efficiently emulating accurate but expensive numerical simulators. Successful applications of such neural simulators can be found in the domains of physics, chemistry, and structural biology, amongst others. Likewise, a neural simulator for cellular dynamics can augment lab experiments and traditional computational methods to enhance our understanding of a cell's interaction with its physical environment. In this work, we propose an autoregressive probabilistic model that can reproduce spatiotemporal dynamics of single cell migration, traditionally simulated with the Cellular Potts model. We observe that standard single-step training methods do not only lead to inconsistent rollout stability, but also fail to accurately capture the stochastic aspects of the dynamics, and we propose training strategies to mitigate these issues. Our evaluation on two proof-of-concept experimental scenarios shows that neural methods have the potential to faithfully simulate stochastic cellular dynamics at least an order of magnitude faster than a state-of-the-art implementation of the Cellular Potts model.