The Darwinian idea of 'survival of the fittest' is central to our understanding of the diversity of life on this planet. However, if only the fittest survive and reproduce, then why do we see so much variation among individuals in traits that are tied to fitness? This problem is especially striking in social systems where cooperating individuals perform some sort of costly act that helps others. Cooperative behaviour therefore has important effects on the fitness of individuals and those that they interact with (often their relatives). Furthermore, cooperating individuals run the risk of invasion by disruptive cheaters that reap the benefits of cooperative behaviours, but do not pay their fair share of the cost. In such situations, we would expect the 'best' strategy to emerge: either cheating or cooperating. Surprisingly, however, studies of natural populations often reveal variation in the degree to which individuals appear to cooperate and cheat. If either cheating or cooperating is the better strategy, then why is there variation along a cooperator-cheater continuum? To better understand this problem, we believe that it is important to not only describe the nature of the variation that is actually present in populations, but also the genes that generate this variation and the processes shaping their variation. This is because, although evolutionary theory may suggest the best strategy, the genetic changes required may not be possible. For example, some strategies may not exist because any gains may be offset by other fitness costs. Alternatively, cooperative traits may be expressed rarely, or there may be limited opportunities to cheat, and as a result the action of Darwinian selection may simply be too inefficient to mould variation to achieve the optimal or favoured strategy. We propose to address this fundamental question using a simple system for the study of cooperative behaviour, the soil dwelling social amoeba Dictyostelium discoideum. Under favourable conditions, D. discoideum amoebae exist as single celled individuals that grow and divide by feeding on bacteria. Upon starvation, however, up to 100,000 amoebae aggregate and cooperate to make a multicellular fruiting body consisting of hardy spores supported by dead stalk cells. Stalk cells thus sacrifice themselves to help the dispersal of spores. Such sacrifices can be favoured because they typically help relatives, but when non-relatives interact, the sacrifices of an individual may help non-relatives. Crucially, like other systems, we have discovered that D. discoideum show enormous diversity in a wide array of traits, including the degree to which different individuals cooperate, thus providing us with a simple system to investigate why such variation exists. To achieve this goal, we will employ a novel combination of approaches in D. discoideum that allow the genetics and evolution of cooperative behaviour and other traits to be analysed with great power. We will use a large panel of naturally occurring strains to identify natural variation in genes that account for the diversity in the traits we observe. We will characterize the types of genes that produce natural diversity in social traits and ask whether those genes also affect other types of non-social traits, which could suggest that they are constrained or shaped by non-social processes. We will be able to determine the types of evolutionary processes that appear to be responsible for the maintenance or persistence of variation in populations. Finally, we will integrate these results with models of evolution to develop a better theoretical understanding of how genetic diversity is maintained and evolutionary outcomes constrained. This work will therefore lead to a fundamental advance in our understanding of the types of variation underlying phenotypic diversity in natural populations and the evolutionary processes shaping that variation.

The transformation of energy in the forms of heat and work pertains to everyday life and is a crucial aspect in the efficiency of machines. In fact, the laws of thermodynamics, which govern these energy transformations, are so fundamental that have their say in almost all branches of physics. The first law acknowledges that heat is energy to be accounted for in energy conservation. The second law of thermodynamics qualitatively distinguishes heat from other forms of energy by associating it to entropy, a measurement of the "lack of information" about a system, and by stating that entropy grows in macroscopic systems. The generality of these statements stems from general statistical properties of macroscopic objects with a large number of degrees of freedoms. However, the technological advances in engineering and operating nanoscale objects like molecular machines, forces us to rethink the implications of thermodynamics for microscopic few-particle systems, where thermal fluctuations are significant. Here the laws of thermodynamics can be reformulated in terms of probabilistic equations, known as fluctuation theorems, which account for rare microscopic events, like those where entropy decreases, which are instead washed away by statistics in the macroscopic word. The formulation and experimental verification of these theorems have been a success of stochastic thermodynamics in the past decade. The nanoscale world, however, challenges us further with quantum mechanical processes emerging at this scale, and devices built upon them. How do we include quantum fluctuations into the laws of thermodynamics? Current research is advancing on this front with some success by analyzing quantum machines operating between classical thermal sources, to identify genuine quantum effects and generalize the definitions of heat and work for quantum processes. The main problem is that in quantum mechanics even measuring the energy of an isolated system is a deterministic process, and that measuring a specified variable, e.g. work along quantum evolution, comes with unavoidable back-action that needs to be taken into account. In this project, we set aside the usual thermodynamic setup where a system is coupled to a thermal bath and focus instead on the measurement process, where a detector monitoring the system is the reservoir with which the system exchanges energy. This kind of configuration allows us to focus on the role of quantum measurement, and it brings new aspects into play, like the fact of dealing with an out-of-equilibrium environment, and the thermodynamic role of the information gained during the measurement. It also comes with the possibility of short-term experimental realizations, since quantum detector's readout is experimentally available, as opposed to thermal baths' readout. The project will set-up the tools to deal with the thermodynamics of quantum measurement and use them to engineer heat flow detectors and possibly heat flow engineering at the nanoscale.

Liquid infused surfaces (LIS) are a novel class of surfaces inspired by nature (pitcher plants) that repel any kind of liquid. LIS are constructed by impregnating rough, porous or textured surfaces with wetting lubricants, thereby conferring them advantageous surface properties including self-cleaning, anti-fouling, and enhanced heat transfer. These functional surfaces have the potential to solve a wide range of societal, environmental and industrial challenges. Examples range from household food waste, where more than 20% is due to packaging and residues; to mitigating heat exchanger fouling, estimated to be responsible for 2.5% of worldwide CO2 emissions. Despite their significant potential, however, to date LIS coatings are not yet viable in practice for the vast majority of applications due to their lack of robustness and durability. At a fundamental level, the presence of the lubricant gives rise to a novel but poorly understood class of wetting phenomena due to the rich interplay between the thin lubricant film dynamics and the macroscopic drop dynamics, such as an effective long-range interaction between droplets and delayed coalescence. It also leads to numerous open challenges unique to LIS, such as performance degradation due to lubricant depletion. Integral to this EPSRC Fellowship project is an innovative numerical approach based on the Lattice Boltzmann method (LBM) to solve the equations of motion for the fluids. A key advantage of LBM is that key coarse-grained molecular information can be incorporated into the description of interfacial phenomena, while remaining computationally tractable to study the macroscopic flow dynamics relevant for LIS. LBM is also highly flexible to account for changes in the interface shape and topology, complex surface geometry, and it is well-suited for high performance computing. The developed simulation framework will be the first that can fully address the complexity of wetting dynamics on LIS, and the code will be made available open source through OpenLB. Harnessing the LBM simulations and supported by experimental data from four project partners, I will provide the much-needed step change in our understanding of LIS. The expected outcomes include: (i) design criteria that minimise lubricant depletion, considered the main weakness of LIS; (ii) new insights into droplet and lubricant meniscus dynamics on LIS across a wide range of lubricant availability and wettability conditions; and (iii) quantitative models for droplet interactions on LIS mediated by the lubricant. These key challenges are shared by the majority, if not all, of LIS applications. Addressing them is the only way forward to better engineer the design of LIS. Finally, the computational tools and fundamental insights developed in the project will be exploited to explore two potentially disruptive technologies based on LIS, which are highly relevant for the energy-water-environment nexus in sustainable development. First, I will investigate application in carbon capture, exploiting how liquids can be immobilised in LIS with a large surface to volume ratio, in collaboration with ExxonMobil. More specifically, liquid amine-based CO2 capture is an important and commercially practised method, but the costly infrastructure and operation prohibit its widespread implementation. Excitingly, LIS may provide a solution to a more economical carbon capture method using liquid amine. Second, motivated by the current gap of 47% in global water supply and demand, as well as environmental pressure to reduce the use of surfactants, I will examine new approaches to clean in collaboration with Procter & Gamble. The key idea is to induce dewetting of unwanted liquid droplets on solid surfaces using a thin film of formulation liquid, thus introducing wettability alteration more locally and using much reduced resources.

Sustainable Development Goals have created impetus to focus on education equity and quality. The education level of millions of children in Low- and Middle-Income Countries (LMICs) out of school for months following pandemic-related lockdowns has dropped. This is particularly troubling considering that, even before the current pandemic, learning outcomes in many of these countries have been poor. Furthermore, crises in LMICs result in multiple negative consequences for disadvantaged children, including a high risk of school dropout, early marriage or pregnancy, displacement, child labour and poor health. Grounded within theories of learning sciences, social justice, human development and capabilities approach, we propose to shift paradigms from a right to education, towards a right to effective inclusion -i.e. equity and quality in learning for all children. This has particular relevance in crisis-affected settings and the context of a pandemic. Adopting an interdisciplinary perspective and using innovative mixed methods, our proposed research builds on a previous intervention, offering a post-COVID-19 lockdowns intervention to make structural and sustainable improvements in education particularly for children from minorities, poor or displaced children, or those with disabilities in rural schools of Afghanistan and Pakistan. Our theory of change offers a pathway towards inclusion, mitigating the negative effects of the COVID-19, climate crisis and political unrest driven school lockdowns through three specific aims: First, we will examine the effects of lockdowns on children's learning outcomes. To measure this impact on child cognitive (e.g. literacy, numeracy, mathematics, logic) and non-cognitive (e.g. critical thinking, communication, self-esteem, coping mechanisms) skills, we will interview, for a new round, children who have participated in an existing randomised control trial to test a similar intervention in 2018, 2020 and 2021. We will also interview teachers and parents to examine: teachers' evaluation of children's skills and knowledge, assessment of their own self-efficacy (including when teaching children with disabilities), perspectives on parental involvement; and parents' socioeconomic backgrounds and satisfaction with education. To understand barriers to learning processes, we will conduct qualitative interviews, focus group discussions, video cues and classroom observations. Second, to remediate for time lost due to the lockdowns, we will further train teachers for two weeks in intervention schools -while engaging with students and parents- in child-centred education practices (Project-Based Learning). We focus on addressing stressors and loss of knowledge due to the lockdowns. We have conducted two rounds of participatory group model building (GMB) workshops under the existing intervention -methods used in Community Based System Dynamics. We will conduct a third round of GMB workshops to engage school stakeholders in highlighting the factors that influence learning in the current situation. Through these workshops, they identify and jointly choose actions to improve the system subsequently put in place with the financial and technical support of our implementing partners -Norwegian Afghanistan Committee, Swedish Committee for Afghanistan, and the National Rural Support Program, Pakistan. Third, we will engage with multiple stakeholders through training, policy engagement and long-term follow-up. Investigators, together with our three partners, are part of the dialogue on education reform in Afghanistan and Pakistan and will further contribute to policy papers based on research findings. They will train development partners (e.g. donors, multilaterals, NGOs) engaged in primary education in both countries. Further, they will collaborate with UNICEF and UNESCO, partner universities and NGOs to disseminate the PBL teacher training module and GMB methods through an e-learning platform.

The aim of this proposal is to develop a novel medical imaging support tool to significantly improve rates of detection, of types of subtle brain abnormality, which give rise to complex brain conditions. Specifically, we are seeking to develop tools that improve the accuracy with which we can compare brain scans across populations. This will make it much easier to tell the difference between healthy and atypical brains, or detect diseased tissue. The reason that this is challenging is because brains are extremely complex, made of billions of cells, and each one can look very different. This makes it hard to build a single model of what "healthy" brains should look like, and as a result it becomes very difficult to spot evidence of disease. These challenges mean that radiologists require years of experience, reviewing countless examples, before they can reliably spot subtle brain abnormalities, and even so, for diseases such as focal childhood epilepsies, up to 30% of cases evade detection. For similar reasons, automated tools often also struggle: appearance of scans varies so extensively that simplifying assumptions must be made leading to coarse solutions. The largest assumption is that all brains share a common organisational blueprint, where areas of the brain responsible for different functions appear in the same order. Such that if each brain scan was a jigsaw, with each piece a region, the shapes might change but they would in go together in the same way. However, in reality brains vary topographically, which means that areas representing different functions (such as language) can swap location. Methods assuming otherwise end up comparing completely different areas of the brain across individuals. Each area may look very different, with different definitions of what is normal. As a result, this leads to confusion, limiting the ability of any method to detect signs of disease. In the past, methods were particularly limited as they built their model of regional organisation based simply on patterns of brain folding. However, it turns out that shape is a fairly coarse and non-specific model of brain organisation, and that brains often have very different patterns of brain folding for the same functional region. Recently we developed a novel open-access tool, which instead learns how to map brains onto a model which takes into account, not just shape but also function, and other aspects of brain organisation (Robinson Neuroimage 2014, 2018). This has led to new, more accurate, models of cortical organisation (Glasser Nature 2016) and development (Garcia PNAS 2018, O'Muircheartaigh Brain 2020) and improved understanding of the links between brain organisation and behaviour (Bijsterbosch Elife 2018). Now we propose to extend this tool, to account for variation of brain shape and appearance in a way that reflects the natural variation seen from one individual to another. Rather than learn a single model of brain organisation we will learn a family of models (modes) that try to describe how our brains vary. These will capture all biologically relevant modes of variation, allowing individual brain scans to be compared, for a given location, only against others with a common organisational blueprint. In this way we will support much more detailed comparison, than was ever possible before. We will validate the power of the approach through three studies: 1) finding the source of epileptic seizures in the brain (to support surgical planning); 2) predicting cognitive outcomes for babies with developmental brain conditions; 3) identifying biological markers in the brain that may help predict mental health conditions. Ultimately, these tools will support researchers, medical doctors and healthcare workers to build more sensitive predictive models, fine tuned to detect signs of abnormality within individual brains. This will improve screening detection rates and lead to more accurate diagnosis of all brain conditions.