Optimal control applied to viral competition
Listen
Radio episode about this paper
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.
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
math-ph, math.MP
Submitted: 2024-11-28
Updated: 2024-11-28
DOI: 10.1016/j.rico.2026.100872
License: http://creativecommons.org/licenses/by/4.0/
Importance score: 81/100
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…
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
Summary
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).
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). p. 4 , to approach
the fixed point of the original system (1), with the less cost possible, therefore and following [37], we
consider the following cost functional, J,
J (VA(t), VB(t), u(t)) = Z tf
0
h
(VA(t) − kA)
2 + (VB(t) − 0)2 + u
2
(t)
i
dt. (4)
This functional is defined in order to approximate the viral density VA to kA and VB to extinction.Where the paper says it
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.
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.
Conclusions
The effect of control means the administration of antivirals influences in two ways First, introducing an appropriate control allows the original system to exchange dynamics Second, choosing the optimal control allows with the lowest-cost treatment to minimize the spread of a more aggressive lineage It is also important to note that the control introduced allows in all cases considered to delay the outburst of the mutant variant and, thus, delay its harmful effects on the hosts.
Declaration of Competing Interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper The acknowledgements section lists several funding sources including the Spanish Ministerio de Economía y Competitividad and European Regional Development Fund.
References
[1] Joshua Lederberg. Infectious history. p. 9 References
[1] Joshua Lederberg. Infectious history.Where the paper says it
[2] Willem G Van Panhuis, John Grefenstette, Su Yon Jung, Nian Shong Chok, Anne Cross, Heather Eng, Bruce Y Lee, Vladimir Zadorozhny, Shawn Brown, Derek Cummings, et al. Contagious diseases in the united states from 1888 to the present, 2013
[3] Jocelyne Piret and Guy Boivin. Pandemics throughout history. p. 9 [3] Jocelyne Piret and Guy Boivin. Pandemics throughout history.Where the paper says it
[4] Fa¨ı¸cal Nda¨ırou, Iv´an Area, Juan J Nieto, and Delfim FM Torres. Mathematical modeling of covid19 transmission dynamics with a case study of wuhan. p. 9 [4] Fa¨ı¸cal Nda¨ırou, Iv´an Area, Juan J Nieto, and Delfim FM Torres. Mathematical modeling of covid19 transmission dynamics with a case study of wuhan.Where the paper says it
[5] Subhas Khajanchi, Kankan Sarkar, Jayanta Mondal, Kottakkaran Sooppy Nisar, and Sayed F Abdelwahab. Mathematical modeling of the covid-19 pandemic with intervention strategies. p. 9 [5] Subhas Khajanchi, Kankan Sarkar, Jayanta Mondal, Kottakkaran Sooppy Nisar, and Sayed F Abdelwahab. Mathematical modeling of the covid-19 pandemic with intervention strategies.Where the paper says it
[6] Nicholas C Grassly and Christophe Fraser. Mathematical models of infectious disease transmission. p. 0 Medical and technological advances allowed scientists to eradicate many
diseases and find treatments for others. Advances in mathematics and the creation of mathematical
models of infectious diseases also made it possible to consider more complex scenarios [6, 7] and prepare
strategies to deal with them.Where the paper says it
[7] Herbert W Hethcote. The mathematics of infectious diseases. p. 0 Medical and technological advances allowed scientists to eradicate many
diseases and find treatments for others. Advances in mathematics and the creation of mathematical
models of infectious diseases also made it possible to consider more complex scenarios [6, 7] and prepare
strategies to deal with them.Where the paper says it
[8] Alejandro Carballosa, Mariamo Mussa-Juane, and Alberto P Munuzuri. Incorporating social opinion in the evolution of an epidemic spread. p. 1 spread and number of deaths. Mathematically, this can be represented by including controls in the
models and solving optimal control problems.Where the paper says it
[9] Garrett Hardin. The competitive exclusion principle: an idea that took a century to be born has implications in ecology, economics, and genetics. p. 0 Therefore, the infected person at the same instant of time is usually affected by
different lineages of the same virus that compete for the same resources. Two different lineages cannot coexist in an individual in the long run as they are competing for the
same resources as follows from the Principle of Competitive Exclusion [9, 10].Where the paper says it
[10] Simon A Levin. Community equilibria and stability, and an extension of the competitive exclusion principle. p. 0 Therefore, the infected person at the same instant of time is usually affected by
different lineages of the same virus that compete for the same resources. Two different lineages cannot coexist in an individual in the long run as they are competing for the
same resources as follows from the Principle of Competitive Exclusion [9, 10].Where the paper says it
[11] Peter J Wangersky. Lotka-volterra population models. p. 1 The competition
parameter, k, in our case, will include other characteristics such as lethality. The simplest scheme for
species competition is based on the Lotka-Volterra equations [27].Where the paper says it
[12] James D Murray. Mathematical biology: I. p. 7 We focused
on the dynamics and the previous analytical study of the system without control and with control, in
order to observe how introducing control allowed us to completely modify the original dynamics. The
described models verify one of the basic principles of biology, the Principle of Competitiv Keywords: Competition model, equilibrium point, optimal control, phase portrait, virus. 1 Introduction
Different diseases and pandemics took place over the centuries or even millennia but only in recent
years have been studied in depth [1, 2, 3]. Elena V´azquez-Cend´on1,3, and
Alberto P. Mu˜nuzuri1,2,∗
1Galician Center for Mathematical Research and Technology (CITMAga), 15782
Santiago de Compostela, Spain
2Group of Nonlinear Physics, Universidade de Santiago de Compostela, 15782
Santiago de Compostela, Spain
3Department of Applied Mathematics, Universidade de Santiago de Compostela,
15782 Santiago de Compostela, Spain
4Department of Mathematics, ISTA, Iscte - Instituto Universit´ario de Lisboa,
1649-026 Lisbon, Portugal
5Center for Research and Development in Mathematics and Applications (CIDMA),
Department of Mathematics, University of Aveiro, 3810-193 Aveiro, Portugal
Abstract
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.Where the paper says it
Where the paper says it
Where the paper says it
[13] Paul Waltman. Competition models in population biology. p. 7 We focused
on the dynamics and the previous analytical study of the system without control and with control, in
order to observe how introducing control allowed us to completely modify the original dynamics. The
described models verify one of the basic principles of biology, the Principle of CompetitivWhere the paper says it
[14] Constantinos I Siettos and Lucia Russo. Mathematical modeling of infectious disease dynamics. p. 9 [4] Fa¨ı¸cal Nda¨ırou, Iv´an Area, Juan J Nieto, and Delfim FM Torres. Mathematical modeling of covid19 transmission dynamics with a case study of wuhan.Where the paper says it
[15] Mirjam Kretzschmar and Jacco Wallinga. Mathematical models in infectious disease epidemiology. p. 0 Medical and technological advances allowed scientists to eradicate many
diseases and find treatments for others. Advances in mathematics and the creation of mathematical
models of infectious diseases also made it possible to consider more complex scenarios [6, 7] and prepare
strategies to deal with them.Where the paper says it
[16] Amit Huppert and Guy Katriel. Mathematical modelling and prediction in infectious disease epidemiology. p. 3 In the following, we develop strategies that allow the
control of the new mutant virus spread by means of the old drugs and treatments. In the framework of virology and epidemiology, it is very useful to apply the optimal control theory
in order to minimize the impact of a disease [35] as well as to minimize the extend of the intervention.Where the paper says it
[17] Eduardo S´aez and Eduardo Gonz´alez-Olivares. Dynamics of a predator-prey model. p. 1 The phenomenon of competition is usually expressed in terms of a predator-prey
mechanism [17, 18]. A clear example is the study of lynx and hare populations [19], but this type
of model is also common in fish species [20].Where the paper says it
[18] Guy Bunin. Ecological communities with lotka-volterra dynamics. p. 1 The competition
parameter, k, in our case, will include other characteristics such as lethality. The simplest scheme for
species competition is based on the Lotka-Volterra equations [27].Where the paper says it
[19] LV Nedorezov. The dynamics of the lynx–hare system: an application of the lotka–volterra model. p. 1 The phenomenon of competition is usually expressed in terms of a predator-prey
mechanism [17, 18]. A clear example is the study of lynx and hare populations [19], but this type
of model is also common in fish species [20].Where the paper says it
[20] Olusola Olabanjo and Ashiribo Wusu. Dynamics of sustainable fisheries: a mathematical approach using lotka-volterra equations. p. 1 The competition
parameter, k, in our case, will include other characteristics such as lethality. The simplest scheme for
species competition is based on the Lotka-Volterra equations [27].Where the paper says it
[21] Javier L´opez-Pedrares, M Elena V´azquez-Cend´on, and Alberto P Mu˜nuzuri. Interactions between hosts affect virus competition mechanism within an infectious strain. p. 0 Two different lineages cannot coexist in an individual in the long run as they are competing for the
same resources as follows from the Principle of Competitive Exclusion [9, 10]. When a mutation occurs
and there are two lineages of the same virus competing within the host organism, the best adapted one
tends to survive in this process of competition resulting in the extinction of the other and verifying
the previous principle.Where the paper says it
[22] Oluwaseun Sharomi and Tufail Malik. Optimal control in epidemiology. p. 3 In the following, we develop strategies that allow the
control of the new mutant virus spread by means of the old drugs and treatments. In the framework of virology and epidemiology, it is very useful to apply the optimal control theory
in order to minimize the impact of a disease [35] as well as to minimize the extend of the intervention.Where the paper says it
[23] Holly Gaff and Elsa Schaefer. Optimal control applied to vaccination and treatmentstrategies for various epidemiological models. p. 0 arXiv:2411. 18998v1 [math-ph] 28 Nov 2024
Optimal control applied to viral competition
Javier L´opez-Pedrares1,2,3, Cristiana J.Where the paper says it
[24] LG Crespo and JQ Sun. Optimal control of populations of competing species. p. 1 Optimal control theory is essential in this type of model, as
it allows the inclusion of interventions like medication to mitigate viral effects [22, 23]. Historically,
in ecosystems, populations were not affected by human impact, except in specific situations such as
hunting.Where the paper says it
[25] Sebastian Sager, Hans Georg Bock, Moritz Diehl, Gerhard Reinelt, and Johannes P Schloder. Numerical methods for optimal control with binary control functions applied to a lotka-volterra type fishing problem. p. 5 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. Again as before, different numerical methods have been used to integrate
the system. Particularly relevant are all the studies done related to the
COVID19 pandemic where the development of strategies that delayed the progress of the sickness where
crucial in saving lives [4, 5]. Medical and technological advances allowed scientists to eradicate many
diseases and find treatments for others.Where the paper says it
Where the paper says it
[26] John P DeLong, Maitham A Al-Sammak, Zeina T Al-Ameeli, David D Dunigan, Kyle F Edwards, Jeffry J Fuhrmann, Jason P Gleghorn, Hanqun Li, Kona Haramoto, Amelia O Harrison, et al. Towards an integrative view of virus phenotypes. p. 0 The
analytical study of these models allows us to obtain an overview of the expected behavior over
time, which is validated with numerical simulations. Keywords: Competition model, equilibrium point, optimal control, phase portrait, virus.Where the paper says it
[30] Yuri A Kuznetsov, Iu A Kuznetsov, and Y Kuznetsov. Elements of applied bifurcation theory, volume 112. p. 0 arXiv:2411. 18998v1 [math-ph] 28 Nov 2024
Optimal control applied to viral competition
Javier L´opez-Pedrares1,2,3, Cristiana J.Where the paper says it
[35] Ali Vahidian Kamyad, Reza Akbari, Ali Akbar Heydari, and Aghileh Heydari. Mathematical modeling of transmission dynamics and optimal control of vaccination and treatment for hepatitis b virus. p. 9 [4] Fa¨ı¸cal Nda¨ırou, Iv´an Area, Juan J Nieto, and Delfim FM Torres. Mathematical modeling of covid19 transmission dynamics with a case study of wuhan.Where the paper says it
[36] JAM Felippe De Souza, Marco Antonio Leonel Caetano, and Takashi Yoneyama. Optimal control theory applied to the anti-viral treatment of aids. p. 7 4 Conclusions
Throughout the manuscript, a simple model of viral competition between two variants of the same
virus, the original strain and its mutation, has been developed and implemented. In this context, we
have simultaneously introduced a treatment for the disease utilizing optimal control theory.Where the paper says it
[37] Aitziber Iba˜nez. Optimal control of the lotka–volterra system: turnpike property and numerical simulations. p. 1 The competition
parameter, k, in our case, will include other characteristics such as lethality. The simplest scheme for
species competition is based on the Lotka-Volterra equations [27].Where the paper says it
[38] Jean-Baptiste Caillau, Olivier Cots, Joseph Gergaud, Pierre Martinon, and Sophia Sed. OptimalControl. p. 5 It is easy to verify that the chosen parameter set verifies the stability conditions. 2 Numerical simulations for the optimal control problem using a direct method
Numerical simulations have been carried out using OptimalControl toolbox [38] implemented in Julia
programming language. 733097 for different initial conditions. 1 Numerical simulations for the (OCP-WSC) using a direct method
As before, using the OptimalControl toolbox implemented in Julia, we can solve the control problem
with constraints on the state variable associated with the mutant lineage.Where the paper says it
Where the paper says it
Green — the paper says it, and the page is printed. Red — no passage carries it, so treat that sentence as unverified. The same check runs on your own text: check a summary against its source, $5.
Improvements for AI systems
-
Adaptive treatment strategy for viral competition: The improved AI system can dynamically adjust antiviral doses based on real-time monitoring of viral strains to minimize cost while maximizing control efficacy, as described by minimizing the cost functional J (4).
-
Predictive modeling of mutant strain dominance: The system can predict which viral lineage will dominate in the absence of intervention by analyzing the fixed points and stability conditions derived from equations (1) and (6).
-
Optimization under state constraints for targeted suppression: The improved system can identify the most cost-effective treatment strategy to keep the mutant strain VB below a critical threshold ξ, as demonstrated by solving Optimal Control Problem with State and Control Constraints (OCP-WSC).
Abstract
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.
Related papers
- Existence of a fluid phase for the two-dimensional one-component plasma
- The Huang-Yang conjecture for the low-density Fermi gas
- Anderson localisation in infinite and finite resonator crystals with long-range interactions
- Free energy and spectral edge of the SYK model
- Anderson localization for 1-d quasi-periodic Schr"odinger operators with degenerate weights
- Irregular SLE(4) martingales and isomonodromic deformations