6 years, 2 babies and a pandemic later, I finally graduated as a Doctor of Philosophy. It was a sunny, winter's day in Oxford, and I got to share the experience with my wife and eldest daughter, it was a real treat. Here is a brief overview of my thesis and a few personal reflections on doing a PhD later in life.

I started my PhD in 2015, at a time when large-scale population genomics was becoming far more accessible, and the question I tried to answer was, how can the study of genome variation within malaria vector populations contribute to the control of malaria in sub-Saharan Africa. The title of my thesis was Genomic epidemiology of malaria vectors in the Anopheles gambiae species complex.

Chapter 1 – An introduction to genomic epidemiology of malaria vectors in the Anopheles gambiae species complex

In the first chapter I gave an introduction to malaria control in sub-Saharan Africa, and the essential role played by mosquito control programmes using insecticide-based interventions. I then introduced whole-genome sequencing and explored ways in which genomics could be used for mosquito surveillance and help to accelerate the development of new mosquito control tools. One of the biggest motivations was of course the knowledge that malaria control has relied heavily on mass distribution of long-lasting insecticidal bednets (LLINs), but Anopheles mosquito populations across Africa have evoloved resistance to the pyrethroid insecticides used in LLINs. Countering this evolutionary threat is a fundamental challenge for malaria control, and genomics could at least help us to gain some visibility of mosquito evolution, giving us a better chance to respond.

figure

Chapter 2 – Historical context: correspondence on the discovery of the Anopheles gambiae species complex

I then took a bit of a detour. The first use of genetics in malaria vector surveillance dates back to the 1960s, when George Davidson and Hugh Paterson used crossing experiments to discover that Anopheles gambiae is a complex of multiple species. I have always enjoyed the history of science, and recounting these discoveries provides some additional background to the results I would present later. But there is another reason for taking this detour, because I have family connections to both Davidson and Paterson. It happens that Hugh Paterson supervised both of my parents' PhDs when they were at the University of Western Australia. When my parents moved to the UK in the 1970s, my father got a job working for George Davidson at the London School of Hygeine and Tropical Medicine. My father stayed in touch with George throughout his life, and after he passed, inherited some of George's possessions. These included a folder containing the letters Davidson and Paterson exchanged on the discovery of the gambiae complex. My father then passed the letters to me, and so I took the opportunity to use them as primary historical sources and explore the story behind the discoveries. I've scanned all of the letters and very much enjoyed being transported back in time to an age when science was different, yet the questions and discoveries were still very relevant.

figure

figure

Chapter 3 – The Anopheles gambiae 1000 Genomes Project phase 1 nucleotide variation data resource

I then got back on track and wrote up all the work I did for phase 1 of the Ag1000G project to build a dataset of nucleotide variation from whole-genome sequencing of 888 individual Anopheles mosquitoes. I analysed the raw variant calls, defined and carried out quality control and validation analyses, investigated genome accessibility and defined variant filters, and produced the final analysis-ready data resource. This chapter is the most technical, but as I learned from doing this work, data quality is vitally important, and there are a number of quality issues that can arise and need to be dealt with.

figure

Chapter 4 – Population structure and genetic diversity

I then explored the Ag1000G phase 1 dataset, in collaboration with other members of the Ag1000G Consortium, investigating population structure and genetic diversity, and how this varies between mosquito populations from different species and countries. One of the most interesting aspects here was discovering how patterns of genetic relatedness are very different depending on which region of the genome you are looking at. This is due to several factors, including large polymorphic inversions and recent selective sweeps. Another interesting discovery was the strong differentiation between mosquito populations on either side of the equatorial rainforest, and the evidence for highly contrasting demographic histories. Clearly the equatorial rainforest has played a major role in shaping Anopheles mosquito population history.

figure

Chapter 5 – Recent positive selection

This chapter focuses on the evidence for recent positive selection in mosquito populations, and the use of genome-wide selection scans to confirm or discover genes involved in insecticide resistance. When I initially ran genome-wide selection scans and shared the results with colleagues, we were all astonished to see the extremely clear and strong signals of selection over some well-known resistance genes. We expected these of course, but we didn't expect the strength or clarity of the signals. In some ways, the high backgroud genetic diversity of mosquito populations helps here, because it creates contrast against which the low diversity created by selective sweeps can be clearly observed. I then worked on a method to automate the detection of selection signals and quantify the strength of evidence, leveraging the fact that recombination breaks down selection signals, giving them a characteristic peak shape.

figure

Chapter 6 – The evolution and spread of target-site resistance to pyrethroid insecticides

In the final technical chapter, I zoomed in on a single gene, which is the binding target of DDT and pyrethroid insecticides, and known to be a locus of insecticide resistance. This was work done in close collaboration with my excellent friend and colleague Chris Clarkson. Although two specific resistance mutations in this gene were known previously, we showed for the first time that each of these mutations has arisen independently multiple times, and has also spread over thousands of kilometres. Mosquito evolution is therefore a formidable opponent, because both mutation and gene flow can contribute to adaptation to insecticides.

figure

Discussion: Towards genomic surveillance systems for malaria vectors

Writing the discussion was a strange experience, because I wrote it at the end of December 2020, just as the newly discovered SARS-CoV-2 "alpha" variant (at the time named lineage B.1.1.7) was spreading through the UK. I lived in Berkshire, but my wife's family lived just across the border in Oxfordshire, and we had been placed under different restrictions. We had planned a big family Christmas, but our plans had to go out the window when the change in restrictions were announced. I wrote:

Although the SARS-CoV-2 virus and Anopheles mosquitoes are fundamentally different forms of life, they share in common the fact that they are both experiencing new selective forces, which are driving their evolution in a way that has major consequences for human health. In the case of SARS-CoV-2, passage from an animal reservoir to human hosts has created a selective pressure to adapt infection pathways and immune evasion mechanisms to their new host’s biology. In the case of malaria vectors, our efforts to eradicate malaria in sub-Saharan Africa through large-scale vector control programmes have created a strong selective pressure to evolve insecticide resistance. However, unlike SARS-CoV-2, we do not have genomic surveillance systems in place for malaria vectors. This means that, as new forms of insecticide resistance emerge and spread within malaria vector populations, we have no means to detect these events, nor to design and coordinate any kind of effective response or mitigation.

The question is, then, how can we now translate Anopheles genomics into a practical surveillance tool.

Reflections

I am not at all one for pomp or ceremony, but I have to say I felt a wonderful sense of completion and achievement at the graduation ceremony today, mixed with a very large dose of relief that it is all finally over! I'm extremely grateful to all the wonderful people who supported me, not least Vikki Simpson who cared about me getting a PhD even more than I did, Martin Donnelly who's idea it was in the first place, and of course my supervisor Dominic Kwiatkowski who provided the inspiration and the opportunity to explore the wonders of mosquito biology.

When I enrolled as a PhD student, I was 37 years old, with one daughter and a full-time job. The logic was that I could keep the job and do the PhD at the same time, because I was already working in a scientific position and so I should be able to re-use a lot of the work I was doing anyway for a thesis. It sounded great in theory, but in practice it was a major challenge. Add to that the arrival of two more daughters along the way, and I really had my work cut out. Making the time to write up was hard when days at work were full and time at home was even fuller with all of the commitments of a young family. It took me six years to complete, and I took every extension I possibly could, stretching the University system, my supervisor, my family and myself to the limit.

In hindsight, part of me does wish I had done a PhD when I was younger and could enjoy the freedom to focus on science with no other commitments or responsibilities. But at the time I completed my bachelor's degree in my early 20s, I just didn't have a passion for doing a PhD in any particular subject, which was why I chose to take a different path and go into computing. The experience I gained has ultimately stood me in good stead, and enabled me to come back into science later with a different set of skills. So perhaps all's well that ends well, and I should embrace the winding road.

If you happen to be in the middle of a PhD and feeling like you have a mountain still to climb, I don't have any great words of wisdom, except to suggest picturing yourself graduating, wearing a fancy gown and hat, receiving a piece of paper and a handshake from someone in an even fancier gown and hat. Imagine how it will feel, because it will feel great, and you can get there!

figure