Optimal control applied to viral competition

summary

The gist

The following manuscript presents mathematical models that describe the emergence of mutant lineages within a viral species and introduces optimal control theory to reduce their growth compared to…

In short

The study develops mathematical models to describe viral competition between an original and a mutant virus using Lotka-Volterra equations. It introduces optimal control theory to determine the most cost-effective antiviral treatment strategy, aiming to minimize the spread of the more aggressive mutant strain while respecting physical constraints on drug dosage.

Key concepts

Lotka-Volterra Equations
These are mathematical models used to describe how two competing populations interact over time. In this context, they model the competition between an original virus lineage and a mutant virus lineage based on their reproduction and competition rates.
Optimal Control Theory
This is a method used to find the best possible sequence of actions (like drug doses) to achieve a desired outcome (like minimizing mutant strain growth) while adhering to certain limits or costs.
Cost Functional J
This mathematical function represents the total cost associated with the control strategy. The goal of optimal control is to minimize this functional, which means finding the treatment plan that results in the lowest overall expense for controlling the virus.
OCP-WSC
Optimal Control Problem with State and Control Constraints. This refers to a more complex optimization problem where not only is the control (drug dose) limited by maximum capacity, but the state of the system (mutant strain level) must also stay within specific boundaries.

Terminology used across episodes

This episode discusses

The paper

Optimal control applied to viral competition · Read on arXiv

Javier López-Pedrares, Cristiana J. Silva, M. Elena Vázquez-Cendγon, Alberto P. Muónuzuri

Galician Center for Mathematical Research and Technology (CITMAga) · Group of Nonlinear Physics, Universidade de Santiago de Compostela · Department of Applied Mathematics, Universidade de Santiago de Compostela · Department of Mathematics, ISTA, Iscte - Instituto Universitário de Lisboa · Center for Research and Development in Mathematics and Applications (CIDMA), Department of Mathematics, University of Aveiro

The emergence of mutant lineages within a viral species has become a public health problem, as the existing treatments and drugs are usually more effective on the original lineages than in the mutant ones. The following manuscript presents mathematical models that describe the emergence of these lineages. In order to reduce the damage and possible casualties that can be attributed to these more contagious microorganisms, the theory of optimal control is introduced and a more sophisticated model is proposed to reduce the mutant growth compared to the original one. The analytical study of these models allows us to obtain an overview of the expected behavior over time, which is validated with numerical simulations.

DOI: 10.1016/j.rico.2026.100872

Transcript

Introduction to the show: ident: Mathematical Physics. The day in mathematical physics: rigorous models of physical systems.

Kai: I'm Kai.

Mira: And I'm Mira, here with Lev.

Kai: We're looking at the paper: "Optimal control applied to viral competition".

Mira: Here's the gist of the paper. The following manuscript presents mathematical models that describe the emergence of mutant lineages within a viral species and introduces optimal control theory to reduce their growth compared to the original one The analytical study of these models allows us to obtain an overview of the expected behavior over time, which is validated with numerical simulations Model Formulation and Competition Dynamics A model of viral competition between two species of the same virus, one from the original lineage and another one from the mutant, can be described by a set of ordinary differential equations The simplest scheme for species competition is based on the Lotka-Volterra equations The model considers two characteristic parameters of the virus phenotype, reproduction rate and competition rate It was shown that viruses tend to adapt with increasing rates of competition while the reproduction rate, r, is not playing any significant role for the system evolution The model is described by equations (1).

Kai: Control Strategy Development Control in the context of virus infections has traditionally been made by humans using drugs and vaccines To account for lineage aggressiveness, two constants cA and cB are introduced to measure the efficacy against viral strain A and B, respectively The control function u(·) represents the antiviral treatment, satisfying the constraint 0 ≤ u(·) ≤ umax ≤ 1 The main goal is to minimize the mutant strain VB(t), in the time window

0, tf: , i.e., to approach the fixed point of the original system (1), with the less cost possible This is approximated by minimizing the cost functional J defined in equation (4).

Mira: Analytical and Numerical Results The analytical study of the problem with constant control allows characterizing the equilibrium points of the system with constant control After calculating the equilibrium points and studying their stability for the problem with constant control, numerical simulations have been carried out in order to observe the behavior for a given set of parameters r, k, c and u As can be seen in Figure 3 for the selection of suitable parameters, with constant control, we are able to invert the dynamics of the original system By applying a constant drug dose, we are able to reach in the new model the equilibrium point equivalent to the unstable one in the original model.

Kai: Impact of Constraints and Optimal Control When state and control constraints are introduced, such as 0 ≤ VB(t) ≤ ξ, a new optimal control problem (OCP-WSC) is formulated The numerical simulations show that the optimal control approach does not achieve eradication but identifies the most cost-effective treatment strategy to control the strains and their anticipated temporal distribution for medication administration In conclusion, introducing a control variable, a treatment, changes the mechanisms of viral competition and makes it possible to obtain lineages that are less harmful to the potential hosts.

Title and authors: Kai: So the title itself tells us a lot about what they're doing here—optimal control applied to viral competition. It suggests they aren't just looking at the dynamics, but finding the best way to intervene in those dynamics.

Mira: Right, and it frames it as an optimization problem where we want to minimize some cost while trying to keep the mutant strain under control.

Lev: Minimizing a cost functional is key; that means they're not just describing what *will* happen, but figuring out what the best action *is* given the system's rules.

Kai: They are aiming to reduce the damage and possible casualties from these more contagious micro-organisms by modeling this intervention.

Mira: The paper sets up a cost functional, which they define as J, and their goal is to minimize that J over a specific time window, say from zero to some time t sub f.

Lev: This whole setup is about finding the optimal antiviral treatment function u of time, where that function has constraints on how much we can actually administer it.

Kai: And they're trying to push the system towards an equilibrium point of the original virus system by using this control input u.

The paper's summary: Mira: To summarize what they do, this manuscript develops mathematical models for viral competition and then introduces optimal control theory to reduce the growth of the mutant strain compared to the original one.

Kai: They start by modeling these strains using Lotka-Volterra equations, which lets them see how they interact based on their respective reproduction and competition rates.

Lev: The models are built around these two characteristic parameters, reproduction rate r and competition rate k, showing how adaptation plays a role in the virus phenotype over time.

Mira: Then they introduce control measures via constants cA and cB to quantify treatment efficacy against strain A and B, which helps them design the intervention function u.

Kai: The main goal is explicitly stated as minimizing the mutant strain VB at time t sub f while keeping that cost as low as possible.

Lev: They use this cost functional J defined in equation four to guide their search for a control strategy that minimizes the impact of the aggressive strain.

The paper's improvements: Kai: The paper suggests a couple of ways they can improve things, first by studying what happens when we allow state and control constraints to be introduced.

Mira: That means they formulate a new optimal control problem, which they call the OCP-WSC, where the mutant strain VB is constrained to stay below some threshold ξ.

Lev: This constraint changes the whole nature of the optimization; it's no longer just minimizing cost freely, but finding a strategy that actively keeps VB within that specific bound.

Kai: The numerical simulations they run show that this approach doesn't achieve complete eradication, but it does find the most cost-effective treatment plan.

Mira: They also look at how this optimal control approach helps identify the best timing and distribution for administering medication over time.

Conclusion: Kai: So, to wrap up, this paper shows that introducing a control variable, a treatment, fundamentally changes the way viral competition works in these models.

Mira: It allows the original system's dynamics to be exchanged or steered in a different direction than it would go on its own.

Lev: And choosing the optimal control means we can minimize spread of more aggressive lineages using the lowest possible cost treatment.

Kai: Crucially, this control mechanism lets them delay the outburst of that mutant variant, which directly delays its harmful effects on potential hosts.

Mira: It’s about using math to guide real-world medical decisions to keep the less harmful lineage from taking over.

More episodes

← Home