Abstract:There has been an increasing recognition of the utility of models of the spatial dynamics of viral spread within tissues. Multicellular models, where cells are represented as discrete regions of space coupled to a virus density surface, are a popular approach to capture these dynamics. Conventionally, such models are simulated by discretising the viral surface and depending on the rate of viral diffusion and other considerations, a finer or coarser discretisation may be used. The impact that this choice may have on the behaviour of the system has not been studied. Here we demonstrate that, if rates of viral diffusion are small, then the choice of spatial discretisation of the viral surface can have quantitative and even qualitative influence on model outputs. We investigate in detail the mechanisms driving these phenomena and discuss the constraints on the design and implementation of multicellular viral dynamics models for different parameter configurations.
What problem does this paper attempt to address?
The problem that this paper attempts to solve is the impact of spatial discretization choices on the system behavior when simulating virus spread in multicellular tissues. Specifically, the author explores how different spatial discretization schemes quantitatively and even qualitatively affect model outputs when the virus diffusion rate is small. By studying in detail the mechanisms behind these phenomena, the author discusses the constraints for designing and implementing multicellular virus - dynamics models under different parameter configurations.
### Background and Motivation of the Paper
Traditional virus - spread models usually assume homogeneously - mixed dynamics and are described using ordinary differential equations (ODEs). However, this assumption ignores the spatial distribution of the virus in tissues. In recent years, more and more studies have begun to focus on spatially - resolved models, which can describe the infection process and its interaction with the immune response in more detail. Spatially - resolved models usually describe the change in virus density through partial differential equations (PDEs) and combine the discrete representation of cells to simulate the spread of the virus in tissues.
### Research Questions
The author points out that although spatially - resolved models have advantages in describing virus spread, relatively few studies have been done on the impact of spatial discretization choices on model behavior. Specifically, the paper focuses on the following questions:
1. **Impact of Spatial Discretization Choices**: How do different spatial discretization schemes affect the quantitative and qualitative outputs of the model?
2. **Mechanism Analysis**: What are the specific mechanisms behind these impacts?
3. **Parameter Dependence**: How do these impacts change under different parameter configurations?
### Methods and Experiments
To answer the above questions, the author conducted a series of experiments and simulations using a typical multicellular virus - dynamics model. This model assumes that cells are distributed on a regular grid, the virus density is described by a PDE, and the virus spreads between cells through linear diffusion.
#### Model Description
- **TIV Model**: The classic virus - dynamics model, which includes the dynamic changes of target cells \( T \), infected cells \( I \), and extracellular virus load \( V \).
- **Multicellular Model**: Cells are represented as discrete spatial regions, the virus density is described by a PDE, and the virus spreads between cells through linear diffusion.
#### Experimental Setup
- **Parameter Setup**: Use biological parameters obtained from the literature, such as infection rate \( \beta \), infected - cell mortality rate \( \delta \), virus production rate \( p \), and virus clearance rate \( c \).
- **Spatial Discretization**: Change the node spacing \( \Delta x \) of the PDE grid to study its impact on model outputs.
### Results
#### Impact of Virus Diffusion
- **Fine - grid Discretization**: As \( \Delta x \) decreases, the time that the virus stays near the source cell decreases, and more viruses diffuse to neighboring cells.
- **Coarse - grid Discretization**: As \( \Delta x \) increases, the time that the virus stays near the source cell increases, but the amount of virus obtained by distant cells increases significantly.
#### Biological Significance
- **Basic Reproduction Number \( R_0 \)**: The choice of spatial discretization affects the calculated basic reproduction number \( R_0 \), thereby affecting the estimate of virus infectivity.
- **Virus - spread Pattern**: Different spatial discretization schemes lead to different virus - spread patterns, especially when the virus diffusion rate is small, this impact is particularly obvious.
### Conclusion
Through detailed experiments and analysis, the author shows the important impact of spatial discretization choices on multicellular virus - dynamics models. These findings provide a theoretical basis for future design and implementation of spatially - resolved models and help to more accurately simulate the process of virus spread in tissues.