Genetic data as a signal of changing malaria transmission

Malaria, an infectious disease spread by mosquitoes, is one of the world's most devastating illnesses. The majority of malaria infections can be found among children living in Africa, where a child dies roughly every minute from malaria. However, in recent years there have been huge advances in malaria control, with the number of yearly deaths due to malaria roughly halving since the year 2000. This is largely the result of increased investment in insecticide treated mosquito nets and effective antimalarial drugs.

As the number of people infected with malaria decreases we face a new challenge - measuring the amount of malaria that remains in a population. Once the majority of the malaria parasites have been eliminated from the population, and the number of people carrying the disease has come down, the remaining cases will tend to be concentrated in a few hard-to-reach groups of individuals. At this point we start to see diminishing returns with ordinary surveillance methods, and it can be difficult to keep track of how much malaria remains. Without accurate measurement we cannot know if antimalarial strategies are working, and it is possible for the disease to bounce back undetected.

One possible way around this problem is to make use of the information held in genetic data. First, individuals suspected of carrying malaria have blood taken and are checked for the malaria parasites (as is usual when diagnosing a case). The DNA of the parasites is also extracted and sent for analysis. Crucially, the level of variation that we observe in the parasite DNA will depend on how much malaria is circulating in the population; if malaria is contained at low levels then we would expect to see very little genetic variation, with all parasites starting to look essentially genetically identical, whereas if malaria is bouncing back to high levels we would expect to see high levels of genetic variation. In this way, genetic data gives us a window into the processes that are occurring in the parasite population without us having to exhaustively sample from every single infected individual in the population.

In my research I will take this relationship between the amount of circulating malaria and the amount of observed genetic variation and turn it into a tool for measuring ongoing levels of transmission. Through mathematical modelling of the malaria life cycle, coupled with modern computational methods, I will design programs that can be used to extract the information present in genetic data, and use this information to estimate the level of infection in a population. Once developed into a user-friendly piece of software, this tool could be used by scientists and policy-makers around the world to make the most of their available data - or to decide whether it is worth the additional cost of collecting new genetic data.

I is my hope that this tool will provide one more line of attack in the struggle to reduce malaria morbidity and mortality globally.

Malaria is declining on a worldwide scale, with many countries now setting their sites on elimination. A major challenge in the future of malaria control and elimination efforts will be monitoring transmission intensity once prevalence reaches low levels, and hence traditional surveillance methods begin to lose statistical power. Genetic data constitutes a rich and underutilised source of information here. The fact that sexual recombination between genetically distinct strains can only occur when an individual receives multiple infections simultaneously (for example by receiving bites from two separate infectious mosquitoes) means that parasite population genetics are intimately linked to ongoing levels of transmission. The handful of studies that have attempted to infer levels of transmission directly from genetic data have shown promising results, but so far lack generalisability and efficiency.

I will develop the statistical and computational machinery needed to turn these fledgeling ideas into a rigorous methodology. By making use of recent developments in coalescent theory, and further extending these methods to account for recombination, I will create efficient algorithms for simulating genetic data from complex malaria transmission models. I will then use approximate Bayesian computation (ABC) to turn this simulation-based method into an inferential tool, using machine learning techniques to maximise statistical power. Finally, I will dedicate time to accessibility, making sure that newly developed methods are available to a wide audience, and not just those used to dealing with mathematical models.

My proposed research has the potential to impact on a number of different sectors:

1. Research scientists

The implementation of my research plan will bring together many methods at the cutting edge of coalescent theory, transmission modelling, and statistical inference. I hope to stimulate interest in each of these areas, and in doing so push malaria modelling to the next level. In particular, I hope to encourage exchange of ideas between the fields of population genetics and pure epidemiological modelling, which have historically remained relatively separate. By producing high quality publications and presenting at conferences in both epidemiology and population genetics I hope to bridge this gap.

2. Malaria policy makers

Monitoring changes in malaria transmission is critical to evaluating the cost-effectiveness of interventions, with surveillance highlighted as a central pillar of the Global Strategy for 2016-2030. Investment in genetic markers is increasingly being promoted as an additional tool to understand malaria receptivity in countries approaching elimination. As new tools become available it is crucial that a framework exists for evaluating their utility, and for making decisions on the basis of firm scientific evidence. My research will help to fill this gap by evaluating the potential for genetic data to enhance existing epidemiological surveillance. Further challenges at the country level include identifying imported malaria cases and preventing the spread of resistance to Artemisinin-based combination therapies. Both of these problems involve gene flow between populations, and so the development of an integrated model of malaria transmission and genetics will be an important first step towards tackling these challenges.

3. Programmes supporting field implementation

One of the specific objectives of my project is the creation of a freely available tool for evaluating the potential benefits of genetic information. This tool will complement the existing Malaria Tools program, which was developed by researchers at Imperial College London in collaboration with the World Health Organisation to aid in malaria elimination scenario planning at the country level. The genetic dimension is currently missing from these tools, and this gap will continue to grow as genetic data becomes cheaper and more easily available. This will make it difficult for field programme directors to know whether it is worth investing in genetic analysis, and if they do intend to invest then what sample size is needed to give informative results. By dealing with this issue at an early stage we can identify those areas where new forms of data are likely to be most beneficial, while also discouraging groups from jumping on the latest technology if it will not prove cost-effective.

4. Future researchers

The rapid progress in genetic technologies in recent years has captured the interest of the non-scientific community, and has the potential to inspire a new generation of researchers. The application of these ideas to problems of health and human welfare adds another dimension to this narrative, which is often lacking from perceived theoretical subjects. By bringing together these ideas in a single project, this research cuts across the fields of biology, medicine and mathematics. Through public engagement activities, teaching and supervision I aim to use this work to inspire future researchers and promote interest in this area.

5. Individuals living in P.falciparum endemic countries

Ultimately the tools developed in my research plan have the potential to benefit individuals living in malaria endemic regions. Through careful monitoring of transmission levels as malaria incidence declines, coupled with informed decision-making and prioritisation of cost-effective strategies, we can ensure that the elimination goals of many countries are reached as quickly and efficiently as possible.