What causes cancer and when?

I am an applied mathematician studying the timing and effect of mutations during cancer evolution. I want to figure out what happens to push benign tumours over the line to become malignant (cancerous). To do this, I am taking a "bottom-up" approach using mathematical models of the underlying mechanisms, and focussing on pre-cancerous conditions and benign tumours.

I'm especially interested in mathematical models of mutational processes that can connect genomic data to population risk and incidence. My goal is to understand how age, radiation and overlapping disorders contribute to tumour risk, and to help discover new genes and mutational processes. My two main foci at the moment are Barrett's oesophagus, a pre-malignant condition; and vestibular schwannoma, a type of benign brain tumour. 

There are three main topics in my current research:

Connecting incidence to genomics

Multi-stage models of cancer evolution have a long history, starting with Armitage and Doll in 1954. The fact that they explain the age-dependence of incidence so well was the earliest evidence that cancers are due to an accumulation of mutations. I am developing mechanistic network models for vestibular schwannoma, to try and link incidence studies to information about specific genes and loci. In these models, different "pathways" can represent different genes, mechanisms, or orders of events. My hope is that this results in new ways to discover genetic risk factors.

Aneuploidy and copy number alterations

An important mechanism that can affect multiple genes simultaneously is loss of heterzygosity (LOH), in which large portions of chromosomes are copied or deleted. This process has historically been simplified away in mathematical models, but I believe it is important to connect models to experiments, and can reveal new mathematical links between genetic diseases and cancer risk.

It is also notable that both cancer incidence and copy number alterations can be described using the mathematics of branching processes. I am interested in developing branching-process-based machine learning models to analyse copy number alteration and other bioinformatic data.

Radiation damage and double-strand break misrepair

An important mechanism of mutagenesis is the induction and subsequent misrepair of double-strand breaks. This can be induced by a dose of radiation, and is often observed in oesophageal cancer. By developing mathematical models of the physical and molecular mechanisms involved, I hope to come up with new insights into the risk of second cancers. I'm also interested in studying protocols and radiosensitisers for radiotherapy.

I completed my Master's in Physics at the University of Exeter in 2013, and went on to write a PhD on biophysics and cancer evolution at the University of Edinburgh, supervised by Bartek Waclaw. I passed my viva voce in 2017. In 2018 I moved to Seattle to take up a post-doctoral position at the University of Washington with Ivana Bozic, working on cancer evolution and models of colorectal cancer incidence. In particular, using evolutionary graph theory to build specific known genes into models of incidence and risk.

I moved back to the UK in late 2019 and worked as a software engineer at the company InSync Technology, which specialised in statistical signal processing.

While in industry, I continued working on my own research when not on company time, applying mathematical models to schwannoma and related diseases in collaboration with Gareth Evans and Miriam Smith. I think schwannoma presents a unique scientific opportunity to relate genomics to large-scale incidence.

In 2022, I took up a position in David Wedge's group, joining the University of Manchester as a Research Associate, researching the genomics of Barrett's oesophagus and other pre-cancerous conditions. In 2023, I won a fellowship from the Congressionally Directed Medical Research Program, to study the effect of radiotherapy on patients with Neurofibromatosis Type-II.

My research interests are the mechanisms that cause low-grade tumours and pre-cancerous conditions to transition to "full" cancers, especially using mathematical models, evolutionary graph theory, and machine learning.

David Wedge

D. Gareth Evans

Miriam J. Smith

• British Neuro-Oncology Society
• Society for Mathematical Biology
• European Society for Mathematical and Theoretical Biology
• American Physical Society

What causes cancer and when?

I am an applied mathematician studying the timing and effect of mutations during cancer evolution. I want to figure out what happens to push benign tumours over the line to become malignant (cancerous). To do this, I am taking a "bottom-up" approach using mathematical models of the underlying mechanisms, and focussing on pre-cancerous conditions and benign tumours.

I'm especially interested in mathematical models of mutational processes that can connect genomic data to population risk and incidence. My goal is to understand how age, radiation and overlapping disorders contribute to tumour risk, and to help discover new genes and mutational processes. My two main foci at the moment are Barrett's oesophagus, a pre-malignant condition; and vestibular schwannoma, a type of benign brain tumour. 

There are three main topics in my current research:

Connecting incidence to genomics

Multi-stage models of cancer evolution have a long history, starting with Armitage and Doll in 1954. The fact that they explain the age-dependence of incidence so well was the earliest evidence that cancers are due to an accumulation of mutations. I am developing mechanistic network models for vestibular schwannoma, to try and link incidence studies to information about specific genes and loci. In these models, different "pathways" can represent different genes, mechanisms, or orders of events. My hope is that this results in new ways to discover genetic risk factors.

Aneuploidy and copy number alterations

An important mechanism that can affect multiple genes simultaneously is loss of heterzygosity (LOH), in which large portions of chromosomes are copied or deleted. This process has historically been simplified away in mathematical models, but I believe it is important to connect models to experiments, and can reveal new mathematical links between genetic diseases and cancer risk.

It is also notable that both cancer incidence and copy number alterations can be described using the mathematics of branching processes. I am interested in developing branching-process-based machine learning models to analyse copy number alteration and other bioinformatic data.

Radiation damage and double-strand break misrepair

An important mechanism of mutagenesis is the induction and subsequent misrepair of double-strand breaks. This can be induced by a dose of radiation, and is often observed in oesophageal cancer. By developing mathematical models of the physical and molecular mechanisms involved, I hope to come up with new insights into the risk of second cancers. I'm also interested in studying protocols and radiosensitisers for radiotherapy.