Multiscale modeling: Difference between revisions

Revision as of 20:19, 25 July 2021

Revision as of 22:05, 7 August 2021

Line 40:

*[[Network theory|Network-based modeling]]

*[[Statistical mechanics|Statistical modeling]]

==Illustrative example: CD4+ T cells==

CD4+ T cells are a key part of the adaptive immune system. They differentiate into different phenotypes to carry out different functions. They do so by secreting molecules called cytokines to regulate other immune cells. In a study, published in 2021<ref>{{Cite journal|last=Wertheim|first=Kenneth Y.|last2=Puniy|first2=Bhanwar Lal|last3=Fleur|first3=Alyssa La|last4=Shah|first4=Ab Rauf|last5=Barberis|first5=Matteo|last6=Helikar|first6=Tomáš|date=2021-08-03|title=A multi-approach and multi-scale platform to model CD4+ T cells responding to infections|url=https://journals.plos.org/ploscompbiol/article?id=10.1371/journal.pcbi.1009209|journal=PLOS Computational Biology|language=en|volume=17|issue=8|pages=e1009209|doi=10.1371/journal.pcbi.1009209|issn=1553-7358}}</ref>, multi-scale modelling was used to explain their emergent behaviors by integrating biological phenomena occurring at different spatial (intracellular, cellular, and systemic), temporal, and organisational scales (signal transduction, gene regulation, metabolism, cellular behaviours, and cytokine transport).

This model uses a Boolean network to describe their intracellular signal transduction and gene regulatory networks, a set of genome-scale constraint-based metabolic models to describe their metabolic reactions, an agent-based model for the cellular behaviours informed by the intracellular models, and compartmental ordinary differential equations to model the reactive transport of cytokines throughout the body. The cells and cytokines reside in and move between three compartments representing the target organ wherein infections occur, the lymphoid tissues connected to the target organ, and the circulatory system.

This is an excellent example of a multi-scale model because it maximises the pros and minimises the cons of each constituent modelling framework. At the systemic level, compartmental ordinary differential equations are suitable since the compartments are considered isotropic and containing indistinguishable molecules. Since the number of molecules is large, they can be solved using deterministic algorithms (relatively efficient). At the cellular level, the agent-based model can consider the heterogeneity in the population and track each cell’s unique combination of attributes. While ordinary differential equations can describe how the sizes of several populations evolve, even cells within one population have different attributes (cell cycle stage, for instance). At the molecular (intracellular) level; signal transduction, gene regulation, and metabolism are considered instantaneous because they are orders of magnitude faster than the cellular events and reactive transport at the systemic level. As such, their transient dynamics are neglected, justifying the use of a Boolean network (for attractor search) and a set of constraint-based models (for optimisation of metabolic fluxes).

In order to facilitate efficient information flow between the four component models spanning multiple scales, a Monte Carlo algorithm was devised. Using the algorithm, the researchers ran a series of stochastic simulations. Through these simulations, they validated the model by reproducing CD4+ T cells' differentiation patterns, metabolic regulation, and population dynamics in response to influenza infection. They finished the study by predicting novel emergent behaviours for this cell type: switch-like activation to filter out non-threatening insults and oscillatory dynamics in response to chronic infections.

==See also==