I am willing to supervise PhD projects in mathematical biology. Please email me if you are interested in the funded project(s) listed below, or in some other topic related to my research; you can find a list of past PhD projects here. Unless stated otherwise below, to apply formally for funding and a place, you must complete an online application form. Please contact me (a.g.fletcher@sheffield.ac.uk) for more information.
Cancer remains one of the most challenging diseases to treat, but recent advances in immuno-oncology (IO) offer new hope. IO therapies aim to use the body’s own immune system to fight cancer. However, many IO clinical trials are unsuccessful, due in part to our lack of understanding how an IO treatment interacts with the immune system and the tumour over time. To improve IO treatments, we need better ways to predict how different therapies will work in patients.
This project aims to develop a more accurate and efficient method for doing just that, by combining state-of-the-art techniques in computational modelling and machine learning.
One key problem in cancer treatment is that every tumour is different, not only between patients but also within a single tumour. The tumour microenvironment, which is the area surrounding the tumour, can have a big impact on how well treatments work. Some parts of a tumour may respond to treatment while others resist, making it difficult to predict the overall effectiveness of a therapy.
Existing computational models that simulate tumour growth and response to IO therapies are often too simplistic to capture these complex interactions, or they are so detailed that they take too long to run on a computer.
This project will develop advanced computational methods to simulate how different IO therapies and drug doses affect both the tumour and the patient’s immune system. By combining two powerful approaches - quantitative systems pharmacology (QSP) and agent-based modelling - we can better represent the complexity of tumour-immune interactions and how they change over time.
QSP helps us understand how drugs behave inside the body at the molecular and cellular levels, while agent-based modelling allows us to simulate the behaviour of individual tumour and immune cells and how they interact in different parts of the tumour.
However, these models can be extremely computationally expensive, meaning they take too long to run for practical use in clinical settings. To solve this, we will use machine learning techniques like multi-fidelity fusion and Bayesian optimization. These are advanced methods that allow us to run simulations more efficiently without losing accuracy. In simple terms, multi-fidelity fusion lets us combine results from quick, less detailed simulations with more detailed ones, balancing speed and precision. Bayesian optimization helps us find the best therapy combinations and dosages by making smart choices about which simulations to run next, saving time and computational resources.
This work will create a powerful new tool for predicting how IO therapies will perform in different patients, building on recent work in this area by the supervisory team at the University of Sheffield and QSP specialists at Certara UK Ltd. It will allow researchers and clinicians to test combinations of therapies and dosages in a virtual setting before moving to actual clinical trials. This could lead to better, more personalized cancer treatments, improving outcomes for patients. The project is suitable for a mathematics, statistics, physics, or computer science student who is keen to learn how mechanistic computational modelling and machine learning can be combined to make progress in biomedicine. The student will be provided with an interdisciplinary training in computational modelling, quantitative analysis and laboratory skills.
A competitively funded studentship is available via the EPSRC Doctoral Landscape Award scheme. Please see this link for information on how to apply. The deadline for applications will be Wednesday 29th January 2025 at 5pm (UK time).
Please contact me (a.g.fletcher@sheffield.ac.uk) for more information.
The evolution of flight in birds is one of the most extraordinary events in natural history. A fundamental event was the formation of feathers, which first appeared in flightless ancestral theropod dinosaurs. In our previous work we showed that the transient inhibition of Sonic hedgehog (Shh) signalling in the chick embryo, led to the failure of flight feather development in the wings of mature birds (Ref 1). Flight feathers are found along the posterior margin of the bird wing and are responsible for most of the flapping, gliding and soaring abilities that enable airborne locomotion. This project builds on our recent unpublished finding that loss of flight feathers causes a ‘polarity shift’ in which the adjacent rows of other feather types are improperly positioned. Based on these preliminary findings we propose a model in which the flight feather buds produce signalling proteins, which initiate the progressive positioning of the other feathers. This involves a process of self-organisation - like the one proposed by Alan Turing in his seminal work - that can be revealed by computational modelling alongside experimental observations.
In this project, we will investigate how feathers are positioned in the chick wing by using a combination of cutting edge embryological/molecular techniques and computational modelling. The findings will have significant implications for the evolution of bird flight and how other structures are precisely positioned during animal development. The project is suitable for a mathematics/physics student who is keen to move into biology, or a biology student with knowledge of cell/developmental biology and genetics who is keen to develop quantitative skills. The student will be provided with an interdisciplinary training in computational modelling, quantitative analysis and laboratory skills.
This project is part of the Yorkshire Bioscience BBSRC Doctoral Training Partnership. Appointed candidates will be fully-funded for 4 years. The deadline for applications will be Monday 6th January 2025.
Lead supervisor: Prof Will Brackenbury Co-supervisors: Dr Andrew Holding and Dr Alexander Fletcher, University of Sheffield
The student will be registered with the Department of Biology (University of York).
Bioelectrical behaviours, such as ion transport and action potential firing, are key features of “excitable” neuronal and muscle cells but are less well understood in “non-excitable” epithelial cells. A key bioelectrical feature of all cells is the voltage difference across the plasma membrane (termed membrane potential, or Vm).
Vm dynamics have significant roles in tissue growth, patterning, migration, and regeneration. Altered Vm may also be a facilitator of disease progression, e.g., in cancer. However, the mechanisms underpinning the Vm, and how it in turn regulates epithelial cell behaviour, are not well understood.
This research, therefore, aims to understand the endogenous bioelectrical mechanisms regulating the Vm in epithelial cells. We will also study how epithelial cells respond to exogenous bioelectrical cues.
To achieve these goals, we will use state-of-the-art electrophysiological and live cell imaging techniques to monitor changes to the Vm in live cells and relate these to alterations in cell behaviour. We will also use cutting-edge optogenetics and chemogenetics to manipulate ion fluxes and perturb the Vm. To understand how epithelial cells respond to exogenous bioelectrical cues, we will develop a multiscale computational model of bioelectrical and metabolic dynamics in an epithelial cell population.
This project will therefore provide novel mechanistic insights into the interplay between exogenous electric fields, ion fluxes, Vm and epithelial cell behaviour. In doing so, the project will expose the student to an array of cutting-edge cell imaging and electrophysiology techniques.
The ideal candidate should have a background in molecular, cell, or developmental biology, and be willing to develop quantitative skills. The student will be provided with an interdisciplinary training in laboratory skills, quantitative analysis, and computational modelling.
This project is part of the Yorkshire Bioscience BBSRC Doctoral Training Partnership. Appointed candidates will be fully-funded for 4 years. The deadline for applications will be Monday 6th January 2025.