-
Configuration from a simulation, with cells to the left by infected state, viral field in the middle, and cytokine field to the right.
Inspiration
The rapidly evolving social and medical impact of COVID-19 requires equally rapid development of tools to predict, and ultimately ameliorate, the future course of the pandemic. Tools are needed to predict patient outcomes, to optimize the deployment of resources by medical providers, and to optimize therapy for infected individuals. Predictive computational simulations of the spread of the disease, and of the underlying biology that makes COVID-19 unique, are both essential. Our aim is to build cooperatively-developed, open-source multiscale model components that can predict various aspects of infection at molecular and tissue scales and inform higher-level models, and a framework to allow others to do so efficiently.
In support of parallel model development of tissue infection, we are developing a multiscale computational model of viral tissue infection and public repository of shared modeling and simulation developments. The repository will be maintained with standard documentation of shared models and simulation tools, for which we will provide supporting templates and generators.
What it does
Currently, the model describes select interactions between generalized epithelial and immune cells and their extracellular environment associated with viral infection and immune response at the cellular and subcellular levels. The repository contains the model and prototypical standard documentation.
How we built it
We are implementing all model components in CompuCell3D, an open-source modeling and simulation environment of the spatiotemporal dynamics of virtual tissue at the cellular and subcellular levels. Model components under development are written in python and simulated in the CompuCell3D simulation engine, which is developed in c++. Supporting tools for documentation standards are being developed in python using Sphinx.
Challenges we ran into
Data on SARS-CoV-2 and COVID-19 is currently limited but increasing by the day. We are filling gaps in knowledge about the virus with data on SARS-CoV, which is much better understood and similar to SARS-CoV-2.
Accomplishments that we're proud of
Currently, the model describes select interactions between generalized epithelial and immune cells and their extracellular environment associated with viral infection and immune response at the cellular and subcellular levels. The framework simulates a dynamic subcellular state of infection in each cell as a system of ordinary differential equations, which is coupled with cellular-level models of cell motility, diffusive environmental viral particles, and cell death due to infection or neighborhood immune cells. Immune cells are recruited by secretion of diffusive signals by infectected cells. The parameter space of the current model includes the outcomes of complete infection and partial infection with residual tissue damage.
What we learned
Spread of infection in tissue is particularly sensitive to the likelihood of viral uptake by individual cells as a function of local viral particles in the environment. Modeling local immune response requires detailed spatiotemporal modeling of cell signaling, including secretion, reaction-diffusion, and chemotaxis models. Local infection modeling in epithelia requires coupling with a model of global immune response for dynamic immune cell populations in simulation as a function of signaling due to the state of infection.
What's next for Collaborative COVID-19 Tissue Infection Model Development
Our immediate goals are to develop specific improvements to the model and shared model repository, including
- Modeling ACE2 in cell uptake of viral particles
- Immune response modeling coupled with the state of infection
- Modeling cell death due to viral exposure
- Define submodel specification for parallel development
Log in or sign up for Devpost to join the conversation.