Abstract
As of today, 20 disease-modifying drugs (DMDs) have been approved for the treatment of relapsing multiple sclerosis (MS) and, based on their efficacy, they can be grouped into moderate-efficacy DMDs and high-efficacy DMDs. The choice of the drug mostly relies on the judgment and experience of neurologists and the evaluation of the therapeutic response can only be obtained by monitoring the clinical and magnetic resonance imaging (MRI) status during follow up. In an era where therapies are focused on personalization, this study aims to develop a modeling infrastructure to predict the evolution of relapsing MS and the response to treatments. We built a computational modeling infrastructure named Universal Immune System Simulator (UISS), which can simulate the main features and dynamics of the immune system activities. We extended UISS to simulate all the underlying MS pathogenesis and its interaction with the host immune system. This simulator is a multi-scale, multi-organ, agent-based simulator with an attached module capable of simulating the dynamics of specific biological pathways at the molecular level. We simulated six MS patients with different relapsing–remitting courses. These patients were characterized based on their age, sex, presence of oligoclonal bands, therapy, and MRI lesion load at the onset. The simulator framework is made freely available and can be used following the links provided in the availability section. Even though the model can be further personalized employing immunological parameters and genetic information, we generated a few simulation scenarios for each patient based on the available data. Among these simulations, it was possible to find the scenarios that realistically matched the real clinical and MRI history. Moreover, for two patients, the simulator anticipated the timing of subsequent relapses, which occurred, suggesting that UISS may have the potential to assist MS specialists in predicting the course of the disease and the response to treatment.
Highlights
Multiple sclerosis (MS) is a heterogeneous disease resulting from the interaction among environmental and genetic factors with a dysregulated immune response
Setting an initial parameter that identifies the random seed of the simulation, the binary strings that are related to major histocompatibility (MHC-I) and II were randomly chosen
We presented a computational framework based on the agent-based modeling (ABM) paradigm that, to our best knowledge, can deliver an accurate description of the disease dynamics, including the effects of the most common treatments
Summary
Multiple sclerosis (MS) is a heterogeneous disease resulting from the interaction among environmental and genetic factors with a dysregulated immune response. Agent-based models represent a simulation technique in which entities, to differential equations systems, are followed individually. In this way, complex emergent behaviors can arise and lead to the prediction of non-coded dynamics. The entities, namely “agents”, are usually placed on simulation space (i.e., a lattice), and in each lattice point, we can have many entities Such agents can be heterogeneous, can have internal properties (i.e., lifetime, internal state, and energy) and can act (i.e., move, interact with other agents in their neighborhood, modify their internal state or die) individually or as a result with the interaction with other agents. Such methods are intrinsically numerically stable as integer numbers represented most of the involved variables, so very few complex floating-point operations are required
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
