Impact of network-structured populations on evolution

Lead Research Organisation: University of Liverpool
Department Name: Mathematical Sciences


Populations are often structured in the sense that individuals are only interacting with their immediate neighbours rather than all individuals. Within these structures, evolutionary dynamics occur. These dynamics underpin many areas including the evolution of our species, the development of antimicrobial resistance and seasonal influenza as well as the emergence of ideas, language and society.

A key way of representing this type of structure is by networks of contacts, and Evolutionary Graph Theory (EGT) has been developed to describe and understand these processes. This structure is idealised in several ways and here, our primary objective is to make this more directly applicable to understanding and describing evolutionary dynamics and especially pathogen evolutionary dynamics.

Evolutionary graph theory assumes a clear distinction between evolutionary dynamics and the underpinning ecological processes of birth and death that drive it. This leads to a lack of realism in the biological processes and limits its utility for addressing real problems. By resolving this, we will obtain a new, more applicable framework. We shall determine the robustness of the theorems of evolutionary graph theory and the extent to which they translate to more realistic scenarios.

Crucially, this model will be the first to be coupled with empirical data from real antimicrobial resistance evolution experiments performed on structured populations in laboratory conditions. This will demonstrate the applicability of the new mathematical framework, providing support for its application in describing real-world systems of pathogen evolution which occurs in structured populations critical for health such as antimicrobial resistance in hospital environments, influenza over airline routes as well as geographic constraints.

Planned Impact

We anticipate longer-term impacts on society through ultimately contributing to shaping healthcare policy. This is through a better understanding of the persistence and drivers of antimicrobial resistance (AMR) and how this is impacted by network-structured populations. In particular, the modelling framework we develop will help us understand evolution of AMR in network-structured populations. We will work with our project partner Public Health England (PHE) to create pathways that will increase the likelihood of our research being used in healthcare policy. One of the longer-term economic impacts we envision is the reduction of disease burden by reducing number of AMR infections in the populations which in turn reduces hospitals costs.

Our combined theoretical-experimental work will focus on AMR in Shigella, which causes shigellosis. Shigellosis is the top bacterial cause of moderate-to-severe diarrhoeal illness in children under five years old in resource-poor settings. The disease kills over 200,000 people per year and the pathogen is resistant to antibiotics, which is particularly concerning for Shigella as no licensed vaccine exists. The lack of management options has led to Shigella being designated a World Health Organisation Priority AMR organism. Many other WHO priority organisms are members of the Enterobacteriaceae family so are closely related to Shigella. This close relationship means our methods should be translatable to other AMR priority organisms.

Other areas in which a better understanding of evolution in network-structured populations should yield benefits include understanding the evolution of strains of seasonal influenza with geographical and transportation constraints as well as the evolution of ideas in social network structures and sexually transmitted diseases across networks of contacts.
Description We have constructed a new mathematical model to represent the impact of network structrued populations on evolution. In particular, this model differs from previous work in enabling both evolutionary dynamics as well as ecological dynamics (of births, deaths and migration) to operate at the same time and thereby consider processes of evolution in which ecological and evolutionary timescales overlap. Better understanding of these processes is important for modelling the evolution of viruses and antimicrobial resistance. Related to the above work, and in combination with laboratory experiments on ecoli, we are developing mathematical descriptions of strain competition dynamics on network-structured populations. These dynamics underpin evolution and we are validating these descriptions against laboratory data.

Related but separate work developed a novel way of determining the effectiveness of the AstraAeneca and Pfizer vaccines on the delta variant of coronavirus by modelling the underlying dynamics of infection against detailed incidence data in the northwest of England. Ongoing work is looking at strain competition dynamics in this data which represents an application of the models we are developing. We are also trying to develop a better undertanding of the evolutionary dynamics of antimicrobial resistance in cystic fibrosis patients using anonymised patient data.
Exploitation Route The model we have developed for evolution could be used to represent real-world systems. As part of the award, we are currently doing this in a laboratory setting for ecoli and also for competing strains of coronavirus and antimicrobial resistance in cystic fibrosis patients.
Sectors Agriculture

Food and Drink



Pharmaceuticals and Medical Biotechnology