Abstract:Many imaging techniques for biological systems -- like fixation of cells coupled with fluorescence microscopy -- provide sharp spatial resolution in reporting locations of individuals at a single moment in time but also destroy the dynamics they intend to capture. These snapshot observations contain no information about individual trajectories, but still encode information about movement and demographic dynamics, especially when combined with a well-motivated biophysical model. The relationship between spatially evolving populations and single-moment representations of their collective locations is well-established with partial differential equations (PDEs) and their inverse problems. However, experimental data is commonly a set of locations whose number is insufficient to approximate a continuous-in-space PDE solution. Here, motivated by popular subcellular imaging data of gene expression, we embrace the stochastic nature of the data and investigate the mathematical foundations of parametrically inferring demographic rates from snapshots of particles undergoing birth, diffusion, and death in a nuclear or cellular domain. Toward inference, we rigorously derive a connection between individual particle paths and their presentation as a Poisson spatial process. Using this framework, we investigate the properties of the resulting inverse problem and study factors that affect quality of inference. One pervasive feature of this experimental regime is the presence of cell-to-cell heterogeneity. Rather than being a hindrance, we show that cell-to-cell geometric heterogeneity can increase the quality of inference on dynamics for certain parameter regimes. Altogether, the results serve as a basis for more detailed investigations of subcellular spatial patterns of RNA molecules and other stochastically evolving populations that can only be observed for single instants in their time evolution.

What problem does this paper attempt to address?

The core problem that this paper attempts to solve is how to infer the stochastic birth, diffusion, and death rates of particles from snapshots of particle positions at a single time point obtained by fixing cells combined with fluorescence microscopy techniques in biological systems. Although this type of snapshot data provides position information with high spatial resolution, it cannot directly reflect the motion trajectories or dynamic changes of particles. The paper solves this problem by establishing a mathematical framework that connects individual particle paths with their manifestations at specific moments (i.e., Poisson spatial processes). Specifically, the paper explores how to use these snapshot data to infer the parameters describing particle dynamics through the method of inverse problems, especially when the number of particles in the data set is insufficient to approximate the solutions of partial differential equations in continuous space. ### Main contributions of the paper 1. **Mathematical foundation**: The paper strictly derives the connection between individual particle paths and their manifestations as Poisson spatial processes, providing a theoretical basis for inferring population dynamics from snapshot data. 2. **Inverse problem research**: Based on the above - mentioned mathematical framework, the paper studies the properties of the resulting inverse problems and analyzes the factors affecting the quality of inference. 3. **Utilization of inter - cellular heterogeneity**: The paper points out that the geometric heterogeneity between cells is not an obstacle but can improve the quality of dynamic inference within certain parameter ranges. 4. **Application prospects**: The results of the paper lay the foundation for more detailed study of the spatial patterns of sub - cellular RNA molecules and other stochastically evolving populations that can only be observed at a single time point. ### Specific methods of the paper - **Model establishment**: The paper assumes that the birth times of particles follow a stationary Poisson point process on the negative half - real line, the initial positions are determined by the source probability distribution, and then move according to the law of Brownian motion, possibly with drift. Particles change states (such as degradation or participation in chemical reactions) at a certain rate, and once they leave the cell domain, they cannot return. - **Statistical inference**: By modeling snapshot data as an inhomogeneous Poisson spatial process of the positions of active particles at time zero, the paper proposes a likelihood function for parameter inference. The form of this likelihood function depends on the way of partitioning the domain, but statistical inference can be optimized by considering the likelihood function in the limit case where the size of the partitioning grid tends to zero. - **Numerical experiments**: The paper verifies the validity of the theoretical results by generating synthetic snapshot data and using Bayesian methods to quantify the joint uncertainty of parameters. ### Conclusion The paper provides a mathematical and statistical framework for inferring particle dynamics from snapshot data at a single time point, which is particularly suitable for the study of the spatial patterns of sub - cellular RNA molecules. By considering the geometric heterogeneity between cells, the paper not only improves the accuracy of inference but also provides theoretical and technical support for future research.

Inferring stochastic rates from heterogeneous snapshots of particle positions

Inferring Stochastic Rates from Heterogeneous Snapshots of Particle Positions

Inference of Spatiotemporal Effects on Cellular State Transitions from Time-Lapse Microscopy

Inferring biological processes with intrinsic noise from cross-sectional data

Inferring Population Dynamics from Single-Cell RNA-sequencing Time Series Data.

The impact of temporal sampling resolution on parameter inference for biological transport models

Incorporating spatial diffusion into models of bursty stochastic transcription

Identifying Drift, Diffusion, and Causal Structure from Temporal Snapshots

Looking forwards and backwards: dynamics and genealogies of locally regulated populations

Modeling of 2D Diffusion Processes Based on Microscopy Data: Parameter Estimation and Practical Identifiability Analysis

Exploiting fast-variables to understand population dynamics and evolution

Modeling Single-Cell Dynamics Using Unbalanced Parameterized Monge Maps

Stochastic force inference via density estimation

Kinetic theories of state- and generation-dependent cell populations

Approximating particle-based clustering dynamics by stochastic PDEs

Accelerating inference for stochastic kinetic models

Accelerated Bayesian inference of gene expression models from snapshots of single-cell transcripts

Computational methods and fast hybrid stochastic simulations for triangulation sensing and identifying principles of cell navigation in the brain

Stochasticity or Noise in Biochemical Reactions

A multiscale compartment-based model of stochastic gene regulatory networks using hitting-time analysis

Inferring Density-Dependent Population Dynamics Mechanisms through Rate Disambiguation for Logistic Birth-Death Processes