UNIVERSITY OF EDINBURGH

UKRI Gateway to Research · FY 2025 · 2025-09

An essential human behaviour is our ability to effectively integrate signals arising from external sensory information (e.g., visual input) with internal cognitive representations (e.g., memory). Such integration allows us to complete seemingly trivial everyday behaviours, such as recalling where you left your keys or imagining your partner's face. However, the mechanisms by which the brain achieves this integration are not fully understood. For example, evidence suggests that perceptual and memory responses might share neural resources in early visual cortex. But, whether this sharing of neural resources extends to other brain regions and different cognitive tasks that also engage externally, and internally orientated processes is less clear. Recent work from our group has shed light on this question by demonstrating the presence of an opponent visuospatial coding scheme in memory-related brain areas. This opponent visuospatial coding was driven by population receptive fields that either responded positively (+ve pRFs) or negatively (-ve pRFs) to a stimulus within their receptive field in a push-pull manner: That is, as activity in one went up, the activity of the other went down and vice-versa. Importantly, at present this opponent visuospatial coding scheme has only been shown to operate within a specific set of brain regions and during a single internally orientated task - cued memory recall. The current proposal aims to test whether opponent visuospatial coding represents a fundamental organising principle of the brain by quantifying the extent to which opponent visuospatial coding structures the brain’s responses during a broad range of externally orientated tasks (controlled and naturalistic visual perception, listening to naturalistic spoken narratives) and internally orientated tasks (cued recall, free recall, speech production and resting-state). We will ask whether opponent visuospatial coding structures the brain's responses across these internally and externally orientated cognitive tasks in both space (where in the brain) and time (when in time). We aim to capitalise on the success of recent precision-fMRI approaches (e.g., Natural Scenes Dataset, [NSD]) that emphasise fewer participants but large volumes of high-quality data. We intend to follow the NSD framework by making these data freely available. To achieve this, we will complete two large-scale projects: During Project 1, we will use cutting-edge functional magnetic resonance imaging (fMRI) techniques to densely sample opponent visuospatial coding in each participant and capture each participant’s responses during tasks that emphasise internal or external cognitive processes. Such an approach will allow us to quantify the strength of opponent visuospatial coding within each participant and the degree to which that coding scheme structures that participant’s responses within the brain (i.e., space). During Project 2, we will use electroencephalography (EEG) recordings during the same internally and externally orientated tasks and in the same participants to capture the temporal dynamics of the brain’s responses. We will then use state-of-the-art computational modelling techniques to quantify the extent to which opponent visuospatial coding structures the brain’s responses across different cognitive states in time. By adopting these complementary approaches this proposal has the potential to identify whether opponent visuospatial coding represents a large-scale fundamental organising principle of the human brain and answer central questions about how the brain integrates internal and external cognitive representations. These are critical issues in Psychology and Neuroscience and fit well within the BBSRC’s Advancing the frontiers of bioscience discovery – understanding the rules of life strategic priority.

UKRI Gateway to Research · FY 2025 · 2025-09

The overall aim of our project is to improve predictions of tropical wetland methane emissions and associated feedbacks in a rapidly-changing climate. We will achieve this by developing and embedding new emission process knowledge, informed by field and laboratory data and by satellite data, into submodels that define the UK Earth System model (UKESM). We focus on Africa for two reasons. First, satellite observations have linked a large fraction of the recent surge in atmospheric methane to wetland emissions from Eastern Africa. Second, Africa exemplifies the compound challenge of accurately describing rainfall, hydrology, vegetation, and the emission of methane, and understanding how they will change in future. To address this challenge, CurFEW necessarily brings together diverse expertise from hydrology, ecology, data assimilation, satellite remote sensing, atmospheric sciences, and wetland dynamics. Previous work by the assembled team has revealed the globally-relevant methane emissions from East African wetlands and documented how they have changed over the past decade. Using satellite observations they have been able to show that a large fraction of observed variations in methane emissions across the tropics is driven by changes in rainfall, driven in turn by large-scale changes in sea surface temperatures (SSTs). Changes in climate driven by rising atmospheric GHGs are expected to cause adjustments to SSTs which led us to propose a new positive climate feedback. Current ESM frameworks are unable to describe these feedbacks because they do not accurately describe the underlying terrestrial processes. In CurFEW, we take a systemic approach to address the key knowledge gaps associated with hydrological dynamics (e.g., increase in runoff, river flow and inundation), and the changes in wetland ecosystems and biogeochemical cycling related to the net release of methane to the atmosphere. These knowledge gaps are relevant to the vast wetland carbon stores across the tropics. Our team comprises world-leading experts that can exploit new satellite, field, and laboratory data to improve how UKESM sub-models of hydrology, vegetation dynamics, and methane emissions will respond to a progressively warmer world. To address our project aim, we have four science questions targeting key uncertainties about how wetland emissions of methane change in a warming world. They necessarily involve integrating models and data on multiple spatial and temporal scales. Q1 What is the role of catchment-scale river hydrodynamics on seasonal and interannual wetland dynamics and on methane emissions? Q2 How strongly does the composition of wetland vegetation (emergent macrophytes such as papyrus and phragmites in perennial wetlands, grasses in seasonal floodplains) control methane emissions? Q3 How will a warming climate impact the future composition, distribution and zonation of wetland vegetation and the associated methane emissions? Q4 At the continental scale, what is the range of future wetland methane emissions under different climate scenarios, and how important are changes in water/land management? CurFEW will deliver a data-driven estimate of future natural methane release from Africa in the presence of climate change and changes in the hydrological cycle. Key research outputs will include: new datasets that lead to insights about the linkages between hydrology, wetland dynamics, vegetation dynamics, and the emissions of methane; improved submodels that define the UKESM; high level peer-review papers; and inputs to national and international assessments. By working with African scientists, we aim to help influence how individual countries address the emerging and contrasting environmental challenges they will face.

UKRI Gateway to Research · FY 2025 · 2025-09

Effective flue gas cleaning to remove sulphur dioxide (SO2) emissions from industrial sources, including Energy-from-Waste (EfW) plants, is crucial for mitigating environmental damage, protecting public health, and meeting stringent regulatory standards. Current technologies must be significantly upgraded to comply with the 2030 National Emission Ceilings Regulations and support the UK’s circular economy goals. The UK government’s comprehensive resource and waste strategy, which aims to eliminate all avoidable waste by 2050, underscores the importance of EfW solutions with efficient flue gas cleaning systems. These solutions reduce the volume of waste sent to landfills, cut emissions, generate energy, and facilitate material recovery, thereby contributing to a closed-loop circular economy. Semi-dry flue gas desulphurisation employs chemical reagents like hydrated lime to absorb SO2 from flue gases. Given stricter emission regulations, it is essential to significantly enhance this process without disproportionately increasing operational costs. Enhanced SO2 removal can facilitate compliance with emission standards through a single-stage flue gas treatment system, significantly lowering capital expenditure for EfW plants and making the technology more accessible, especially for developing countries. Efficient SO2 removal with reduced reagent consumption will decrease production and delivery costs of hydrated lime, thus reducing transportation carbon emissions and making the overall operation more economical and sustainable. Moreover, it will minimise the production of residues that need treatment or landfill disposal. Minimising residue production also lowers the fouling risks and the need for plant shutdowns for manual cleaning, thereby increasing plant availability and potentially saving costs. Achieving SO2 removal enhancements requires moving beyond trial-and-error methods and developing a robust theoretical foundation for process design and optimisation. A prerequisite for further improvements is a fundamental understanding of the underlying physics of two common semi-dry technologies: Particle-Powder Spouted Beds and Circulating Fluidised Bed Reactors. Both technologies involve interactions between solid particles (e.g., hydrated lime), liquid droplets (e.g., water or slurry), and gas phases (flue gas). Factors such as temperature, humidity, particle size, and particle spatial distribution significantly influence these interactions, making numerical modelling and accurate SO2 removal predictions challenging. Furthermore, physical models characterising the interfacial transport and chemical processes require further experimentally informed development. Fulfilling this modelling gap is the main aim of the present proposal. We will take a micro-to-macro scale approach to comprehend the SO2 removal process in semi-dry desulphurisation, which involves (i) understanding the effects of operating conditions, (ii) exploring the effects of complex particle interactions, and (iii) developing online monitoring techniques to characterise inter-particle dynamics and quantitative SO2 removal rate simultaneously. Non-uniform temperature and velocity distributions in large-scale desulphurisation reactors cause particles to experience different localised conditions based on their position. To address this, we will develop a parametric regime map for micro-scale transport phenomena by conducting single-particle experiments in an acoustic levitator under high temperature and humidity conditions. The complex interactions between solid particles and liquid droplets in semi-dry desulphurisation reactors introduce significant hydrodynamic challenges. We will investigate inter-particle interactions in an innovative vertical wind tunnel with a diverging cross-section to develop experimentally informed models. Furthermore, the impact of inter-particle behaviour on overall bed performance and SO2 removal is not well understood. We will adapt and utilise the experimental methodology of the Depth-from-Defocus to evaluate particle spatial distribution and inter-particle distances and combine it with quantitative SO2 measurements using Planar Laser-Induced Fluorescence to create an online monitoring system.

UKRI Gateway to Research · FY 2025 · 2025-09

This project will enhance the agrigenetic simulation ecosystem to supercharge theoretical and applied studies in selective breeding and the adoption of their results in practice. Similar to other branches of science, simulations are pivotal to research in modern data-driven selective breeding of agricultural populations. These simulations are also an essential tool for decision making in applied breeding; on how to deploy new data generation processes and associated data analysis methods to increase the efficiency and return on investment. These uses of simulation have a significant academic, industrial, and societal impact by advancing the fundamental role of agriculture – food production. There is now a vibrant and complementary ecosystem of high-quality simulation software packages. While some of the software packages are already interoperable, further developments are required to increase interoperability and to fully leverage their distinct strengths. Specifically, we need three key innovations to improve the agrigenetic simulation ecosystem. First, we need to increase interoperability between population genomics and selective breeding simulation software packages by leveraging the succinct tree sequence encoding of ancestral recombination graphs, including support for the common diploid genomes and complex polyploid genomes. Second, we need to increase flexibility of agrigenetic simulation software to generate complex phenotypes as a function of a range of effects with user-defined relationships to account for the nature of agriculturally important traits and concepts such as adaptability, resilience and genotype-by-environment interactions. Third, we need to increase the use of vast amounts of publicly available genomic data in an easy-to-use and affordable way. This project will deliver these needs through three work packages: i) Supercharging genetic and selective breeding simulations with tree sequences, ii) Simulating complex phenotypes that capture environmental variation and its interaction, and iii) Accessible genomic and phenotypic data resources for agriculture. Through these work packages we will extend the software packages AlphaSimR and msprime to increase interoperability of genomic data and add support for complex polyploid genomes, all via tree sequences. We will further develop a flexible PhenoSimR package to create a general framework for simulating complex phenotypes. Finally, we will infer tree sequences from publicly available genome-wide genotype and whole-genome sequence data of six key agricultural species (wheat, potato, maize, cattle, sheep, and pig). We will host these tree sequences on a server and create deployable server containers to offer an accessible live demonstration of the developed tools from this project. This project will enhance a popular agrigenetic simulation ecosystem and will further boost its uptake in applied settings and academia. Project development was guided by significant community demand. The work will be undertaken with the advice of an expert board and supported by our project partners. This community is collectively delivering significant academic outputs and a majority of the UK and world-wide genetic gains through selective breeding of their populations. There is an urgent need for the development of the proposed tools due to the growing world population and climate change, both increasing pressure on agriculture to produce more food with greater efficiency and sustainability. The cutting-edge results from this project will expand the ecosystem of agrigenetic simulations to contribute to solving these challenges with data-driven approaches. The project will also promote use of publicly available genomic data that has to date been underutilised. These results provide the essential springboard to develop future digital twins for selective breeding programmes.

UKRI Gateway to Research · FY 2025 · 2025-09

Too many children and adolescents are dying globally, with particularly high numbers of deaths in Eastern and Southern Africa. To ensure policymakers can make informed choices on how to prioritise interventions and optimise health systems to reduce mortality requires data on the causes of death in children and adolescents and an understanding of how the risk of these varies across sociodemographic groups. While in most high-income countries there are robust systems to capture cause of death information on the whole population, these do not generally exist in low-income countries. In the absence of such systems, demographic surveillance sites (DSS) have been established to collect demographic data at regular intervals within geographically defined populations (e.g., on the number of births and deaths in the population) (Herbst et al, 2021). In DSS, reported deaths are typically followed up with a verbal autopsy (VA), an interview where a relative or caregiver provides information on the signs and symptoms, and the social and healthcare circumstances, preceding the death which can be used to assign the most likely cause of death (Chandramohan et al, 2021; Fottrell and Byass, 2010). Unfortunately, the accuracy of cause of death assignment using VA remains particularly poorly understood among children and adolescents. There are two main approaches to assigning cause of death based on VA information. The first method is physician review, where 2-3 physicians review the VA and assign cause of death. More recently, however, algorithms have been developed to automate the cause of death assignment (Byass et al, 2019; McCormick et al, 2016; Serina et al, 2015). Our aim is to provide population-based estimates of the causes of child and adolescent deaths (0-19 years old) in Eastern and Southern Africa, and assess how these vary by sociodemographic characteristics. To do this, we will use a rich VA dataset that has been compiled by the Analysing Longitudinal Population-based HIV/AIDS data on Africa (ALPHA) network, harmonising data from seven DSS across five countries in Eastern and Southern Africa and covering the period 1994 to 2022. The specific objectives are as follows: To assess the quality of VA data for children and adolescents. To compare methods of assigning cause of death for children and adolescents from VA data. To provide population-based estimates of the causes of child and adolescent death, and look at how these vary by key sociodemographic characteristics.  To explore the social and health system circumstances surrounding child and adolescent deaths. There are several important applications anticipated. Firstly, we will provide recommendations on using algorithms to assign cause of death in children and adolescents from VA. For example, we will ascertain at what level of detail causes can be reliably assigned (e.g., where we can attribute a death to specific cause of death, such as “malaria”, and where we can we only ascertain cause of death at a broader level, such as “infectious disease”). Secondly, we will provide training to DSS teams on how to conduct these analyses, so there is a strong platform to produce similar estimates at regular intervals in the future. Finally, we will provide much needed empirical estimates of the sociodemographic and healthcare drivers and causes of child and adolescent mortality in a region lacking these data, which can be used by country policymakers as well as international agencies such as the World Health Organization.

UKRI Gateway to Research · FY 2025 · 2025-09

The chemical sector emits 26 Mt CO2 annually and 95% of its inputs are currently fossil-derived. Engineering biology holds vast potential to improve the sustainability of this growing sector, yet faster progress is required to achieve the ambitious Net Zero targets. To address this, CIRCBIONET (Circular Bioeconomy through Engineering Biology Network for Specialty Chemicals) aims to establish a world-class Engineering Biology (EngBio) consortium to catalyse the specialty chemicals industry's transition to sustainable, biosynthetic processes which are embedded within the circular economy by design. Leveraging unique expertise from across the team, this will be enabled by integrating specialty chemical bioproduction with valorisation of end-of-life plastics using a biotechnological ‘upcycling’ approach. This goal will be met through four interconnected objectives: To develop novel biosynthetic pathways for three high-value specialty chemicals: capsaicin, folic acid and rhamnolipids. To develop capability to use a new non-conventional microbial host, Bacillus subtilis, for engineering biology pipelines. To valorise waste polymer-derived feedstocks for specialty chemical production which will promote the future circular economy. To embed automation, metrology and workflow standardization across all work packages to generate robust datasets for future machine learning, life cycle assessment and techno-economic analyses projects.   To achieve the objectives, seven interconnected work packages are proposed. The activities span the design of biosynthetic pathways for conversion of enzymatic polymer degradation products to specialty chemicals, identification of new degradation products from enzymatically processed polymer waste, and engineering of a novel microbial host to assimilate waste-derived degradation products. In addition, data-driven bioprocess optimization will be conducted using one of the engineered microbes to produce one specialty chemical. As an underlying enabling technology, the programme will embed automation, metrology and standardization of workflows into the work packages for integration into life-cycle assessments, techno-economic analysis and machine learning in the future. A major innovation in CIRCBIONET is the use of deconstructed, post-consumer polymeric materials, such as poly(ethylene terephthalate) (PET), poly(carbonate) and poly(urethane), as feedstocks. The choice of specialty chemicals (capsaicin, folic acid, rhamnolipids) and biodegraded polymer wastes (PET, PU, PC) as feedstocks have been chosen to combine existing capabilities of the members. These exemplar projects will enable knowledge transfer between EngBio experts in the UK and Singapore to accelerate innovations and build the foundations for future collaborations, benefitting both countries. Another unique feature is the use of Bacillus subtilis as a host for the conversion of polymer waste into specialty chemicals. Attempts to engineer microbial hosts for conversion of PET monomers have been reported mostly on E. coli which is Gram- and only one attempt on Gram+ bacteria (Rhodococcus jostii). B. subtilis is a model Gram+ bacteria that is of industrial relevance for its ability to synthesize molecules, to secrete proteins, and to utilize multiple carbon sources. This will be the first systematic exploration on B. subtilis for specialty chemical production from waste polymers and will expand its utility as a chassis in EngBio. In summary, the proposed CIRCBIONET will contribute to generation of specialty chemicals through biorecovery of waste materials by unlocking embedded carbon and keeping it in circulation for longer. The data-driven approach along with automated metrology and workflow standardizations that will be integrated into the programme will bring engineering biology closer towards Industry 4.0. CIRCBIONET will facilitate knowledge transfer between UK and Singapore spanning seven leading institutions for achieving sustainable development goals and circular bioeconomy.

UKRI Gateway to Research · FY 2025 · 2025-09

Future-proof secure communication enabled by quantum cryptography offers proven resilience and security across various industrial sectors, making it of utmost national importance. The UK National Quantum Strategy envisions achieving quantum advantage at scale by 2035, reaching a trillion quantum operations, which could undermine the classical cryptographic techniques currently used to secure the internet. This underscores the significance of quantum-key distribution (QKD) in achieving provable security in the presence of future quantum computing-enabled attackers. The existing limitations on the key generation rate, reliability, and cost of the current QKD technology necessitates faster, more reliable, and more practical QKD solutions. QUANTA’s vision is to produce a significant step change in the design and performance of practical QKD systems. In contrast to current QKD receivers, which use individual single-photon detectors (SPDs) to measure qubits, QUANTA will employ SPD arrays to exploit their scalable spatial degrees of freedom. This enables QUANTA's receivers to support additional functionalities beyond qubit detection, such as wavefront sensing, spatial beam tracking, multiplexing, noise cancellation, and dead-time mitigation. Thus, a range of advancements can be achieved from increasing resilience to the turbulence effect in atmospheric QKD links deployed on terrestrial or satellite platforms to enhance the key generation rate of low-cost wireless QKD systems.

UKRI Gateway to Research · FY 2025 · 2025-08

The Standard Model of particle physics is our best theoretical knowledge of the Universe at the shortest length scales ever probed in experimental measurements. However, there is a number of things it does not seem to explain, such as the puzzle why there is observable matter in the Universe. Such shortcomings of the model can indicate the existence of new, hitherto unknown particles and forces that would have to be searched for in future measurements. In this project, I will test the limits of the Standard Model by performing timely and ambitious calculations, which when combined with high-precision experiments allow to indirectly search for new physics. I will focus on composite particles known as hadrons as a potential portal to the unknown, through (A) scattering of pions, (B) the muon magnetic moment and (C) decays of kaons. To handle effects from the strong nuclear force (QCD) in the hadrons, the work will rely on the complementary tools effective field theory and lattice QCD. In (A) I will develop a framework to include electromagnetism and sub-leading QCD corrections in lattice QCD simulations of pion scattering. These effects are typically expected to be of the percent level size, and therefore must be taken into account for any precision goal at (sub-)percent level. The main goal here is to obtain a relation between the pion energy levels in the lattice simulation and physical scattering parameters, known as a quantisation condition. The developed formalism will solve a fundamental problem for lattice simulations, and set the stage for a wide range of future calculations.  In (B) I will scrutinise a questioned tension between theory and experiment for the muon anomalous magnetic moment. I will consider one of the two most important hadronic contributions, the so-called light-by-light currently known at 20 percent relative precision. In dispersion theory evaluations of the light-by-light, parts of the calculations rely on models. Here I will use fundamental constraints from QCD derived by myself and collaborators in recent years, needed to reach the 10 percent precision goal on the light-by-light. In (C) I will use the quantisation condition from pion scattering in (A) to lay out a lattice QCD strategy for decay of kaons into two pions including electromagnetism. This decay is particularly interesting as it is related to the matter asymmetry observed in the Universe, and it is special since the sizeable electromagnetic effects are expected to be at the 20 percent level. The work will entail finding a relation between what can be obtained in lattice QCD and the physical decay, known as a Lellouch-Lüscher factor. This kind of factor is typically needed for hadronic decays, and the one obtained here will be the first to include also electromagnetic effects. The results will therefore open for a wide variety of future lattice simulations. This heavily impactful project exploits the complementarity between effective field theory and lattice QCD. The host organisation is The University of Edinburgh, which has a long tradition of numerical expertise and thus nicely complements my own background. The work will be successful as I have extensive background with the needed ingredients, thus making the novel results highly feasible. Without the expected results in (A) and (C) it will not be possible to do Standard Model precision tests including scattering effects in lattice QCD at (sub-)percent precision.

UKRI Gateway to Research · FY 2025 · 2025-08

Multiple myeloma is the second most common type of blood cancer, affecting approximately 6,000 new cases in the UK every year. This malignancy originates from plasma cells, a type of immune cell, which undergo genetic changes that trigger uncontrolled cell division, leading to significant health complications. Our genetic code is organised into strings of DNA within every cell, where the active units, known as genes, serve as the fundamental instructions guiding cellular function and behaviour. Each gene occupies a designated position within one of these strings, ensuring coordinated biological processes. However, in the case of myeloma cells, this precise arrangement can become disrupted; like a string breaking and tangling, causing genes to get lost, duplicated, or rearranged. The exact breaks and affected genes differ from one patient to another. Such variability in genetic alterations underlies the wide range of symptoms and treatment responses observed in myeloma, presenting a major challenge in understanding and treating this cancer. In an initial study of 23 myeloma genomes, we discovered something unexpected. Over 60% of these cases had DNA breakages in areas that are usually tightly packed and protected from such damage. These areas, known as compact heterochromatin, had been a mystery until the latest human genome map in 2022 revealed their DNA sequence, enabling them to be studied for the first time. Yet, extracting meaningful information about these areas through genome sequencing presents significant analytical challenges, leaving them largely unexplored in any cancer. To address these challenges, we have developed a new method that enables us to uncover the intricate details and implications of these breakages. Our findings indicate that these breakages are not random but are associated with worse outcomes for patients, affecting crucial genes that can aid tumour growth. Our project aims to dive deeper into these genetic abnormalities and study their significance in myeloma. We suspect that disturbances in these compact DNA regions could lead to a cascading effect, making the myeloma cells more robust and resistant to treatment. By analysing a large set of myeloma genomes and employing modern genetic techniques across various molecular levels, we aim to pinpoint the architecture and mechanisms of these changes. We will identify their precise locations and the genes involved, exploring their influence on myeloma development from its asymptomatic stage through to advanced disease progression and treatment resistance. The insights from this study promise to revolutionize our understanding of multiple myeloma, identifying key factors that could predict the disease's biology and uncover new ways to target the cancer cells. Furthermore, the computational methods we develop will have broad applications in cancer research, offering new avenues to study genetic abnormalities across different cancers.

UKRI Gateway to Research · FY 2025 · 2025-08

Sensitive free-text data - such as clinical notes, educational information, and social care records - captures rich, narrative information essential to understanding societal issues, health inequalities, and social context. This kind of unstructured data offers insights often unavailable in coded fields. However, its use in research remains severely limited due to the unpredictable presence of personally identifying and potentially sensitive information. As the UK's research infrastructure increasingly adopts a federated approach to securely analyse sensitive data across institutions, the challenge of using free-text safely and responsibly has become more pressing. Although policy frameworks for the safe use of structured data in Trusted Research Environments are well established, no equivalent standards exist for provisioning free-text for secure research use. This gap leads to uncertainty for data custodians and overly cautious risk assessments, ultimately restricting access to data. Free-text Challenge: Unlike most structured data, identifiers in free-text are embedded in context and vary by language use, document type, population, and institutional practice. They are often difficult to detect using rule-based methods, especially when rare or implicit (e.g., mentions of suicide, rare diseases, or recognisable local places). Current de-identification approaches, which are typically trained on small, restricted datasets, do not generalise well across diverse sources or settings. Moreover, the lack of auditability and transparency in existing tools raises concerns for both governance teams and the public. Large language models (LLMs) offer a promise to better understand the nuances and context specific nature of identifiers in unstructured text, but due to restrictions there is very little evidence of the benefits they bring on real-world sensitive data. Our aim in this project is to advance the responsible reuse of clinical free-text in federated research environments by building scalable, LLM-enabled privacy-risk solutions; enhance previous (DARE UK SARA) privacy-risk tools for cohort-level privacy decision-making; and generate evidence and insights shaped by community and governance to augment the Standard Architecture For Trusted Research Environments (SATRE) guidelines to include free-text use. Objectives: 1.     Evaluate and develop scalable LLM-enabled tools for de-identifying sensitive free-text. 2.     Enhance our prototype privacy-risk management and audit tool for cohort-level governance decisions. 3.     Generate real-world insights to inform safe, community-aligned use of free-text in federated research and strengthen SATRE guidance. Critically, our work is shaped through ongoing engagement with public contributors, data custodians, researchers, and governance professionals to ensure that automation supports, not replaces, human oversight. Working with public contributors and the TRE community, we will generate a reproducible, evidence-based output to guide the safe use of free-text across UK research environments, ensuring its value can be realised without compromising privacy, public trust, or ethical standards. Working with the TRE community we will use these outputs to enhance the SATRE guidelines where applicable to include free-text use. This project tackles an important gap by developing both new technology and strong oversight to help researchers use sensitive free-text data safely and responsibly at scale.    

UKRI Gateway to Research · FY 2025 · 2025-08

Viruses are a major threat to human health and wellbeing. The cellular antiviral response represents the first line of defence against viruses. Characterising intrinsic cellular defences is necessary for the development of new broad-acting antiviral treatments. Central to the cellular antiviral response is the generalised shutdown of protein production, or translation, within the cell. Viruses are entirely reliant on cellular protein production machinery. Generalised translational shutdown in an infected cell helps limit viral replication and spread. However, at the same time, cells must produce a host of antiviral defence proteins that inhibit viral activity directly. Balancing these two contradictory processes, while also limiting viral manipulation of protein production, can determine the outcome of viral infection. While the regulatory mechanisms that drive translational shutdown are well understood, far less is known about those allowing concurrent selective translation of antiviral proteins. In this research study, I will examine how changes in the cellular translational machinery itself could drive the selective production of defensive antiviral proteins during host translational shutdown. I will focus on the core components of the translational machinery: ribosomes, the large macromolecular machines that carry out protein synthesis, and tRNAs, the adaptor molecules that decode genetic instructions into protein. Changes in the cellular tRNA pool and ribosomal populations can drive the expression of specific proteins in various contexts, including cancer metastasis and embryogenesis. Moreover, viruses and infected host cells have been shown to manipulate tRNAs and ribosomes to affect the outcome of infection. However, we do not have a comprehensive characterisation of the extent and impact of changes to the tRNA pool and ribosomal population during the antiviral response. I hypothesise that cells specialize their tRNA pool and ribosomal population to enable selective translation of defensive proteins during the antiviral response. To test this hypothesis, I will use tRNA sequencing, ribosome profiling, and mass spectrometry-based protein identification to generate a high-resolution picture of the tRNA pool and ribosomal population over the course of the antiviral response in human cells. I will identify significant features of this antiviral translational machinery and characterise their functional relevance in the selective translation of host defence proteins. This project will provide insight into a key component of the antiviral response and will elucidate mechanisms by which regulation of the translational machinery can affect cell fate and function. Understanding what drives selective translation during the antiviral response will improve our ability to therapeutically bolster cellular antiviral defences and impede viral manipulation of cellular protein production.

UKRI Gateway to Research · FY 2025 · 2025-08

Bioimaging is one of the fastest evolving technologies. Just a decade ago visualising and measuring developmental and disease processes in molecular detail in living cells, in three dimensions, was unimaginable. While standard confocal and wide-field fluorescent microscopes can image in 2D, bioimaging in the third dimension in live samples over time is a challenge. High laser light doses damage biological processes and kill cells and organisms. Slow image acquisition and distortions in the Z (or axial) plane limit 3D studies. A new generation of microscope which use sheets of light for 3D live imaging has been pioneered to address these limitations. We are applying for the most recent version of this technology a Lattice Light Sheet Microscope. This new instrument uses structured sheets of light to speed image acquisition, enabling fast acquisition of 3D images at sub-micron resolution with minimal distortion. Zeiss launched the Lattice Light Sheet 7 (LLSM) in 2022, developed in consultation with core facilities and end users. It is straight-forwards to use, does not require complicated sample preparation, is stable over several days and able to image a wide range of samples using micron thin sheets of light. We need sub-micron resolution for our science. Previous iterations of the technology couldn't show us the detail needed due to: lack of resolution, sample preparation requirements which were incompatible with our biomedical research applications, issues with the illumination field or complexity of operation for our facility users. The equipment will be housed in the Advanced Imaging Resource (AIR), Institute of Genetics and Cancer (IGC), University of Edinburgh (UoE). AIR comprises equipment and six expert technical staff have particular expertise in very high-resolution imaging of fixed and live samples. Coupled with methods developed from our molecular and developmental biologists, this will ensure that the LLSM is used in a wide array of biomedical applications. The IGC has specialist expertise in interrogating how small changes to DNA in the genome impact development, cancer and disease. We are delighted to have several early career researchers onboard with this application. The LLSM will be used to tackle biomedical questions including; How gene regulation from the non-coding genome controls development of the face and eye,  How variations in the human genome contribute to poor cancer, and other disease, prognosis and outcomes How liver bile ducts form and are repaired How the molecular construction of cilia contributes to disease We are committed to open science. To ensure the technology is available to a wide audience we will incorporate the LLSM into ESRIC (The Edinburgh Super-Resolution Interdisciplinary Consortium), a national centre of excellence for super-resolution microscopy. ESRIC is a collaboration between the IGC and Heriot Watt University, it offers an open access core facilities across both sites, with some equipment being housed within AIR - 'a facility in a facility'. ESRIC is one of seven sites which comprise the UK node of EuroBioimaging, providing national infrastructure for open-access state-of-the-art bioimaging resources. We offer specialist training courses to pass on our expertise. To summarise, the LLSM will enable a significant step forward in research capabilities, at a large MRC Unit, within the IGC, regionally within Edinburgh. It will also benefit research both nationally / internationally through the EuroBioimaging UKnode.

UKRI Gateway to Research · FY 2025 · 2025-08

A fundamental component of human social cognition is our ability to reason about emotions. Reasoning about our own and others’ emotions is essential to building and maintaining social relationships that are critical to health and well-being throughout the lifespan – in humans and across species. Despite its importance, key open questions limit our understanding of emotion reasoning and development in human children and, by extension, the potential benefits of this research for people and societies.   Specifically, the normative developmental trajectory of emotion understanding in school-aged children (i.e., 5–10-year-olds) remains unclear because children’s performance on emotion tasks improves at the same time as continued language development. This makes it difficult to know whether children’s understanding of emotions undergoes development during this time, or if apparent development actually reflects improved language abilities. As a consequence, there is disagreement about children’s emotion understanding and development, which has implications for how educators, parents, and clinicians support children to develop positive, healthy relationships, and help children to identify and reason about their own and others’ emotions and social experiences. Research investigating links between emotion understanding and real-world outcomes (e.g., development of social relationships, mental well-being) also often suffers from this confound – making it difficult to develop strategies to support children’s social and emotional development.   The proposed research aims to address this challenge through the use of non-invasive brain imaging (functional magnetic resonance imaging; fMRI) studies of adults and children. When adults and children reason about emotions, a particular network of brain regions is recruited; this network is sometimes referred to as the ‘mentalising’ network. While the mentalising network interacts with nearby brain regions that support language, it is specifically recruited to reason about mental states (i.e., beliefs, desires, emotions), across linguistic and non-linguistic contexts. Putative homologous regions in non-human primates suggest some evolutionary continuity in this system and its role in supporting social interactions across species.   Because the mentalising network is recruited for emotion reasoning and largely insensitive to linguistic features, we can use fMRI to tease apart contributions of domain-specific change in emotion understanding to developmental improvements on emotion reasoning tasks among children. If developmental improvements on emotion tasks involve change in emotion understanding, then behavioural improvements should be accompanied by developmental change in the mentalising network. Our studies will also characterise concurrent development in language brain regions, to provide a full picture of the interactions between mentalising and language systems in human development.   The proposed research leverages recent advances in adult fMRI research; specifically, we will use ‘multivariate’ analyses to capture fine-grained response patterns in the mentalising network to test specific, competing theoretical accounts of children’s emotion understanding and development. We will also use these data to characterise the links between mentalising network development and development of social relationships and mental well-being – to understand the nature of these relationships early in human development and inform education, clinical, and parenting practices.   This research will provide critical insights into children’s emotion reasoning and development, establish an approach for future research studying other aspects of cognitive development in children, and forge new links between emotion research and real-world social behaviour and outcomes. More broadly, this research will significantly accelerate our understanding of a fundamental component of human social cognition that supports health and well-being throughout the lifespan.

UKRI Gateway to Research · FY 2025 · 2025-08

Geometric phase is a unifying concept that spans both classical and quantum physics. Pancharatnam--Berry (PB) phase is an optical manifestation of geometric phase which provides a measure of the dissimilarity of two electromagnetic plane waves of the same frequency, based on the evolution of their polarization states. Currently PB phase is the focus of intense research activity, especially involving anisotropic and/or non-homogeneous materials in which its effects are most striking. Applications of PB phase in wavefront tailoring, polarization-dependent lenses, holograms and waveguiding are being pursued. These developments have been fuelled by advances in engineered materials, including metamaterials and metasurfaces. Voigt waves propagate in bulk materials. Voigt surface waves propagate at the planar interfaces of materials. Both are singular forms of optical plane-wave propagation that arise when the matrix governing their propagation fails to exhibit Hermitian symmetry and has repeated eigenvalues. In contrast, non-singular propagation is characterized by Hermitian symmetry and distinct eigenvalues. A distinguishing feature of Voigt waves and Voigt surface waves is that both are unusually localized. Voigt waves decay in the direction of propagation in a manner that depends on the product of the propagation distance and its exponential (in contrast to non-singular plane waves whose decay is purely exponential). Voigt surface waves decay in the direction perpendicular to the planar interface in a manner that depends on the product of the distance from the interface and its exponential (in contrast to non-singular surface waves whose decay is purely exponential). This unusual localization may be exploited for applications in optical sensing and communications. Material anisotropy is essential for the non-Hermitian conditions to be satisfied for both Voigt wave and Voigt surface wave propagation. Crucially, Voigt wave propagation and Voigt surface wave propagation are both polarization dependent. Indeed, materials supporting Voigt wave propagation can discriminate between left- and right-circularly polarized light. To date research on PB phase has concentrated exclusively on non-singular optical propagation -- the issue of PB phase for singular optical propagation has not been considered. In particular, an open question is as yet unanswered: What PB phase is associated with Voigt waves and Voigt surface waves? The proposed research will address this matter by calculations of PB phase for plane waves transmitted through slabs of anisotropic materials. The transfer-matrix method, which is a powerful mathematical formalism that has been developed to determine the reflection and transmission characteristics of slabs of complex materials, will be applied to calculate the PB phase. Since this method can be adapted to accommodate singular optical propagation for the most general types of linear materials, it is ideally suited to the proposed research. Specifically, the PB phase will be calculated for the transmitted wave relative to the incident wave for: (i) Voigt-wave propagation within an anisotropic slab;  and (ii) Voigt-surface-wave propagation at the planar interface of an anisotropic slab. The PB phase will be related to the symmetries and constitutive parameters of the supporting anisotropic materials. By choosing the supporting anisotropic materials to be engineered materials, such as homogenized composite materials, the constitutive parameters of the supporting anisotropic materials may be varied to tune the PB phase as required. Therefore, not only will our understanding of PB phase for singular optical propagation be elucidated, but a means of tailoring the PB phase to suit the requirements of particular applications will be developed.

UKRI Gateway to Research · FY 2025 · 2025-08

Cleaning is a necessary chore in all homes. Whilst the pandemic has firmly associated cleaning with surface disinfecting; cleaning– as practiced in the home environment– is a collection of different tasks that interact with the three-dimensional space. In many specialist environments such as commercial food preparation and healthcare, cleanability is a regulated design requirement. Yet, in the domestic setting, cleanability is not a quality that most architectural or interior designers strive for. Indeed, the impact of design on cleaning is not well understood. However mundane, cleaning takes time, such that the amount of time spent doing it is an indicator of quality of life, especially for women – who still carry out the majority of housework in the UK and shoulder increased cleaning chores since COVID-19. Market research indicated that almost half of the UK adults felt that keeping the home clean is stressful. Yet not all households can afford to outsource cleaning, which is especially sensitive to the cost-of-living crisis. The ability to perform basic home-cleaning is also a requirement for older people (who cannot afford cleaning services) to age-in-place and live independently. Still, for the most part, the narrative of cleaning in the domestic sphere is dominated by products and know-hows, not the design characteristics of the physical space itself. This project aims to examine and elucidate how the design of our homes impacts their cleanability. Specifically, two key questions are explored. First, what attributes – beyond the time required to perform the task – make up the quality of “cleanability” in the home environment? This line of enquiry examines the household cleaning experience as a person-environment transaction where the process (the experience) is valued as much as the end result (getting the cleaning done). Second, what design characteristics positively or negatively contribute to the cleaning experience? How do these characteristics vary by room functions or occupant needs? As a pilot focusing on the UK context, this project tackles these questions first through a series of semi-structured interviews with professional domestic cleaners in the Midlothian region who have experiences cleaning different housing types. Then, a number of households in Edinburgh will be recruited to share their cleaning experiences via a walk-through of their homes and cleaning routines. A subset of resident participants will be shadowed during one of their cleaning routines where difficulties in their cleaning experiences can be observed and mapped to design details. This project has two objectives: (1)To investigate and develop the concept of “cleanability” based on household cleaning experiences; and (2)To construct a taxonomy of design characteristics that contribute to (or hinder) the cleanability of domestic spaces. As the first investigation into the relationship between design characteristics of domestic spaces and residents’ cleaning experience, this project will contribute to better design guidelines that can help modify or create living spaces that reduce the burden of cleaning and home maintenance, and thereby support ageing-in-place and make the familiar chore a more efficient and enjoyable experience for all who do their own cleaning and are time-poor.

UKRI Gateway to Research · FY 2025 · 2025-08

Improving healthspan is a key goal for society. Underlying this is a drive to understand why some people appear to age faster than others. If this can be identified and quantified then we can examine if different organs and tissues age at the same rate and test if modifiable lifestyle behaviours and the external environment impact this biological ageing. Over the last decade, we have driven research to show that DNA methylation (DNAm, an epigenetic modification) patterns from blood are leading candidates to assess biological ageing. Our pioneering work has linked DNAm with chronological ageing as well as modifiable lifestyle behaviours and traits linked to healthspan e.g., smoking, alcohol consumption and BMI. In addition to providing a readout of an individual’s environment, we have shown that DNAm, which can be thought of as analogous to a dimmer switch to turn gene expression up/down, is partly regulated by our underlying DNA sequence. Despite these advancements, numerous challenges remain. These include understanding differences in DNAm patterns by tissue/biosample type. Almost all large DNAm studies have focused on blood methylation. Given the proximity of the oral cavity to the external environment and the possibility of remote, inexpensive and non-invasive sampling, saliva has potential as a biomarker of healthy ageing. By understanding the genetic regulation of salivary DNAm, we can identify if differences in patterns are caused by lifestyle and environmental stimuli and test if this is also the case for blood-based DNAm. A second limitation of existing studies is how longitudinal patterns (e.g., DNAm changes with age) are typically inferred from cross-sectional data in studies with wide age ranges. Repeat sampling from the same individuals requires following study participants over many years, which is a costly process. However, within-person trajectories represent the gold-standard and enable us to examine if changes in biology mirror changes in lifestyle behaviours. Finally, while a host of disease-based DNAm studies have been conducted, few have sought to take a complementary approach to identify patterns that characterise healthy ageing. With multimorbidity (the accumulation of multiple disease diagnoses in an individual) being an increasingly observed phenomenon, understanding biological patterns that promote disease-free ageing are crucial. Here, we will address the limitations in previous research by analysing the largest blood- and saliva-based DNAm datasets in the world in tandem with longitudinal blood-based DNAm and concurrently measured blood- and salivary-DNAm. By integrating these data with detailed measures of the social, cognitive and physical environment in an outstanding research environment, we will: 1) determine the genetic patterns that underlie the salivary methylome and compare their similarities to the genetic regulation of the blood methylome. We will use these findings to ascertain if environmental and behavioural patterns cause differential DNAm patterns at overlapping or different genes across blood and saliva; 2) discover which genes and regions of the salivary methylome show differential patterns by age and modifiable lifestyle factors and; 3) identify methylation-based signatures of healthy ageing and investigate in longitudinal samples if these change over time and in tandem with markers of physical and cognitive health. We will develop and apply a suite of novel statistical data analysis tools to address these points. Understanding how the genetic regulation and lifestyle correlates of the methylome vary by tissue will provide fundamental insights into the role of DNAm for tracking both biological and healthy ageing.

UKRI Gateway to Research · FY 2025 · 2025-08

As urban areas become more densely populated, effectively managing and predicting pedestrian dynamics is critical in areas such as public safety, architecture and transportation. For example, crowd managers could quickly assess different escape routes during an evacuation to minimise injuries; architectural design can be optimised by considering the dynamics of pedestrian movement; and the development of autonomous vehicles could benefit from a better understanding of how people navigate crowded urban spaces. However, current simulation tools lack the ability to model large, dense crowds in real time. On the one hand, the agent-based approach, in which pedestrians are modelled individually, is so computationally expensive that it makes real-time crowd management impossible. On the other hand, current models that liken crowds to continuum 'thinking fluids' are severely limited in their accuracy because, in order to obtain a closed system of equations, they typically make simplistic assumptions about constitutive relationships, such as that relating the local speed of pedestrians at a point to the state of the crowd in the area around that point. This proposal will derive the first fluid dynamics-like crowd model using a data-driven modelling technique to combine the scalability of the continuum description with the accuracy of data-driven approaches. As the opaque 'black box' models provided by most machine learning methods are not suitable for high stakes applications such as those involving crowds, a novel machine learning approach will be developed that extends the current ability to distil mathematical expressions from the data. We will use two complementary data sources - firstly, surrogate pedestrian models, as these provide clean datasets from controlled experiments that are ideal for developing the machine learning tools, and secondly, real crowd observation data to ensure the accuracy of the distilled model. Integrating this model into OpenFOAM (a widely used open-source computational fluid dynamics platform) will create the first high-fidelity simulator of large, dense crowds in real time. The simulator is expected to significantly reduce the time required for crowd management planning, optimise architectural design for better pedestrian flow, and improve the navigation of autonomous vehicles in crowded urban areas, ultimately reducing the risk of crowd related accidents and injuries, which sadly increase every year (an updated list can be found here). The applicant will lead a research team consisting of a postdoctoral research associate (funded by this grant) and a PhD student (funded by the University of Edinburgh). An experienced academic partner will act as a mentor, ensuring effective project management and providing expertise in the application of machine learning tools to fluid dynamics. Two industrial partners will also be involved in the project: Buro Happold's expertise in pedestrian flow analysis will be integrated into the machine learning tools, while ESI-OpenCFD's expertise in fluid dynamics will facilitate the seamless integration of the simulator into OpenFOAM. The results of the project will be disseminated through high impact papers, conference presentations, public engagement at festivals and the open-source release of the simulator. FLOCKS is expected to have a significant impact on the events industry by reducing the time required for crowd management planning and the percentage of crowd related accidents or injuries. The high level of interest shown by the industry partners in their letters of support underlines the importance of the project and their commitment to its success.

UKRI Gateway to Research · FY 2025 · 2025-08

Pre-clinical research relies on in vitro and rodent studies. Unfortunately, despite enormous investment, between 86–95% of drugs fail to show efficacy or gain approval for use. The reasons for these failures are complex, but the tendency to jump from rodent derived therapeutics into the clinic (often via pharmacodynamic and toxicity studies in non-human primates) likely plays a role. The use of large animal models, which better replicate human anatomy/physiology and underlying disease pathophysiology, has been used successfully to address this issue. Whilst this may be somewhat contentious, to improve translational hit rates, we need to integrate pre-clinical large animal models into the drug development process. Advanced, clinically relevant, imaging techniques play a vital role in validating large animal models of human disease and in evaluating new treatments. The University of Edinburgh’s (UoEs) Large Animal Research and Imaging Facility (LARIF)?is a world-leading research facility; opened in 2021, it represents the culmination of a £24M investment. LARIF is dedicated to large animal translational and livestock research, and is well-equipped to translate medical breakthroughs up the clinical path for human patients. Alongside state-of-the-art surgical suites, LARIF has imaging modalities comparable with a human hospital, including optical coherence tomography, ultrasound and a 3 T MRI platform. However, we currently rely on a now ageing Siemens SOMATOM Definition AS64 CT scanner. State-of-the-art 16 years ago and originally supported by the Wellcome Trust to establish our critical care laboratories, the scanner has been instrumental in many of our studies of human disorders in large animal models. These studies have included pesticide toxicity, neurodegenerative conditions, medical device development for pulmonary and cardiology applications and, most notably, a £10M research project modelling human lung tumours in sheep. We have a wealth of expertise and workflows attracting researchers to our facility and now seek to further bolster our standing in this space, leveraging from existing investment, to procure a new dual source CT scanner with increased image resolution, dynamic imaging capabilities and dedicated in-room interventional capabilities. These advancements would result in improved image quality, shorter scan times, and reduced subject and operator radiation exposure, all of which are requisites for conducting advanced biomedical studies. To reinforce LARIF’s position at the forefront of biomedical/translational research, this machine would focus on modelling interventional procedures and precision/personalised medicine approaches, comparable to that being performed in human clinics. Our overall objective is to maintain LARIFs status as a BBSRC National Bioscience Research Infrastructure, enhance imaging capabilities and attract new academic/industrial partners. This will allow us to develop new research areas, including cohort studies to understand the pathophysiology of diseases, experimental medicine studies to assess potential therapeutic interventions and even preclinical randomised controlled trials aimed at informing human trials to improve patient outcomes. Having imaging facilities engineered for a range of clinical situations also provides opportunities for healthcare practitioners to undertake procedural development and training. This application sits at the heart of MRCs strategic delivery plan, in particular their investment into the development of medical imaging within the areas of neuroscience, cancer, cardiovascular and respiratory systems. Altogether, this application has the ability to reinforce the Royal (Dick) School of Veterinary Studies pre-eminent status and the University’s world ranking, thereby helping fulfil UKRI’s role in delivering the government’s ambitions for the UK as a global leader in research and innovation.

UKRI Gateway to Research · FY 2025 · 2025-07

Proteins are the workhorses of life; they are the molecules that carrying out most of the essential functions inside our cells that keep us alive. Many of these proteins don't work alone; they associate together with other proteins and small molecules (such as vitamins and lipids) in intricate partnerships called protein complexes. Additionally, the biological function of protein complexes is often tightly controlled by the presence or absence of specific partners within the complex or chemical modifications on key proteins. To fully understand how life works at the molecular level we need scientific techniques to study protein complexes to: Identify the individual members within a complex. Understand how these complexes are assembled and how different members of the complex interact with each other. Determine how the activity of a complex is controlled and regulated.  Mass Spectrometry (MS) is one analytical technique which has proven to be a powerful tool for addressing these challenges. New technology advances in the last few years now mean that the latest MS instruments can study larger protein complexes with more speed and sensitivity as well as gain more detailed information about the chemical modifications of individual proteins within complexes. Here we invest in the latest-generation mass spectrometer at the University of Edinburgh. The enhanced features of this new instrument will enable us to perform experiments that have previously been impossible, enabling us to break boundaries in biological research. Our team comprises research technical professionals (RTPs) as well as academics/researchers from all career stages. Our RTPs are expert in using mass spectrometry to create new approaches to study biological molecules and have the required technical skills and experience to realise the full potential from the new capabilities of this next-generation instrument. The mass spectrometer will be housed in the School of Chemistry’s Scottish Resource Centre for Advanced Mass Spectrometry (SIRCAMS), a facility that has been at the forefront of biological mass spectrometry for two decades. Researchers will work closely with RTPs to apply this new technology to help understand how proteins and protein complexes work in areas which are key strategic priorities for BBSRC: e.g. improving our understanding of health and healthy aging, e.g. by investigating the formations of large protein complexes that occur in neurons and how these contribute to neurodegenerative diseases such as Alzheimer’s and Parkinson’s Disease. developing innovation in renewable resources and clean growth, e.g. to study proteins engineered to include unnatural building blocks that can perform chemical reactions not seen in nature, as an alternative to traditional less sustainable approaches.   advancing research in sustainable agriculture and food production, e.g. to study and engineering the key protein complex in plants that performs photosynthesis in order to increase crop productivity. After installation, we will make this new technology available to the wider research communities, e.g. other academics institutions in Scotland and the North of England as well as partners from industry, ensuring maximum wide-reaching national impact. In implementing this project, we will train and mentor our RTPs in advanced technical skills and support and educate the next generation of bioscience researchers.

UKRI Gateway to Research · FY 2025 · 2025-07

The availability of very large and high-dimensional data sets poses great challenges to decision makers in many settings such as medical diagnosis, loan applications, fraud detection, and so on, each of which increasingly relies on machine learning models for automated decision making. The identification of a small number of relevant features of the data to make accurate predictions not only significantly reduces the complexity of the decision, but also dramatically enhances its interpretability, transparency, and fairness, all of which play a fundamental role in explainable artificial intelligence. Such decision problems can be mathematically modelled as optimization problems with cardinality constraints (OPCC), i.e., optimization problems with an upper bound on the number of nonzero components of the decision variable (also known as cardinality, sparsity, or l0-norm). Sparsity plays a vital role in a wide range of applications such as machine learning, data science, signal and image processing, portfolio management, sparse regression, and compressed sensing due to the interpretability, robustness, and ease of implementation of sparse solutions. As such, optimization problems with cardinality constraints arise in a plethora of applications, such as feature selection in machine learning, sparse support vector machines, sparse principal component analysis, and sparse portfolio optimization. Since the l0-norm is a discontinuous, integer-valued, nonconvex, and nonsmooth function, OPCC constitutes a highly challenging class of optimization problems. In pursuit of computational tractability, previous research has focussed on heuristic approaches based on tractable convex surrogates such as the widely popular l1-regularisation. However, this popular approach is only guaranteed to recover an exact solution of OPCC under either very restrictive or difficult-to-verify assumptions. Furthermore, for general OPCC, they can lead to highly inaccurate and grossly misleading solutions. The pervasiveness of OPCC in many critical applications, such as medical decision making, and the drawbacks of popular heuristic approaches necessitate research efforts towards finding optimal or near-optimal feasible solutions. With the notable exception of the sparse portfolio optimization problem, the literature on exact or approximate solution methods for OPCC is in its infancy. The main goal of this project is to close this gap by studying exact and approximate solution methods for OPCC in a unified framework. We aim to achieve this goal through two objectives: (1) Exact Solution Methods: We aim to perform a rigorous comparison of exact formulations of OPCC in terms of robustness, scalability, and optimality gaps considering different modelling options and the use of special types of constraints. (2) Approximate Solution Methods via Convex Relaxations: Since exact solution methods are not scalable, we aim to develop fast and effective feasibility restoration methods for constructing near-optimal feasible solutions of (OPCC) using optimal solutions of various tractable convex relaxations and rigorously analyse the trade-off between solution quality and scalability. This project will lay the foundations for finding optimal or near-optimal solutions for this class of problems in a unified framework. Improved solution accuracy and reliability will have significant benefits for a host of other fields that employ machine learning, artificial intelligence, data science, and business analytics models.

UKRI Gateway to Research · FY 2025 · 2025-07

Context and research the equipment will enable The axolotl is a unique vertebrate model for biomedical research. Axolotls are best known for their ability to regenerate virtually any body part and organ, including limbs, tail, lung, heart, thymus, kidney, large parts of the brain and spinal cord. Less well known is that some early developmental processes such as germ cell specification are conserved between axolotl and mammals. Additionally, axolotls age relatively slowly and rarely develop cancer. These features highlight the potential of the axolotl to provide information relevant to human health, in addition to their obvious relevance to developmental, evolutionary and regenerative biology. The mechanisms that underlie the axolotl’s astonishing regenerative capacity and intriguing biology remain poorly understood. Axolotls are an excellent model for developmental biology, as their eggs are large and embryos develop externally. The recent sequencing of the axolotl genome and the emergence of new technologies such as single-cell multi-omics and CRISPR/Cas9-based genome editing, already successfully employed in axolotl, now make it possible to gain deep mechanistic understanding of axolotl development, regeneration, slow ageing and cancer resistance. We envision that this fundamental understanding, and the resulting comprehensive comparative analyses between regenerative and non-regenerative species, will inform new strategies to improve human health. Axolotls are clearly poised for discovery. However, we lack axolotl research capability in the UK. Aims and objectives We aim to build an axolotl research pipeline in the Institute for Regeneration and Repair (IRR), at The University of Edinburgh (UoE). This will include axolotl breeding and experimental rooms to become the only axolotl research colony in the UK. The axolotl facility will enable a wide range of new research avenues, from regeneration, development, and stem cell research, to ageing, cancer and evolutionary biology; and spark new collaborations locally, nationally, and internationally. Potential applications and benefits The facility will be integrated within IRR, a joint venture between the School of Biological Sciences (SBS) and the College of Medicine and Veterinary Medicine (CMVM) to form the largest grouping of scientists and clinician scientists investigating tissue repair and regeneration worldwide. The critical mass of researchers at IRR and the wider UoE means we are uniquely positioned to reveal fundamental differences between highly regenerative axolotls and poorly regenerative mammals, including humans. This will allow us to leverage key advances in regenerative biology relevant to regenerative medicine. There is widespread enthusiasm for axolotl research among developmental and regenerative biology communities across the UK. However, new UK import regulations (following Brexit) and the fact that axolotls are critically endangered in the wild and listed under Appendix II of the Convention on International Trade in Endangered Species (CITES), make it extremely difficult for UK researchers to access axolotls or axolotl tissue from abroad. Building an axolotl research pipeline in the UK is therefore essential to boost biomedical research and attract world-class scientists. Additionally, new collaborations will be fostered by allocation of flexible space for visiting researchers within the facility and through annual open workshops. The axolotl facility will be managed by Bioresearch and Veterinary Services, which oversees UoE's animals facilities. This will offer unique career development and research opportunities on axolotl welfare to this team of research technicians and technology and skills specialists (RTPs). Our proposed axolotl research pipeline will deliver new fundamental knowledge with transformative potential for regenerative medicine and human health.

UKRI Gateway to Research · FY 2025 · 2025-07

The IMPACT imaging facility delivers microscope imaging capability for over 300 scientists, students and support staff within the Centre for Discovery Brain Sciences (CDBS), Institute of Neuroscience and Cardiovascular Research (INCR), University of Edinburgh (UoE). As part of the Edinburgh Bio-Imaging Consortium we also offer our services across the whole UoE and to external academic and commercial clients. Our confocal microscopes are currently being used at near 100% of capacity (where 100% capacity is considered to be 1400 hours per year), each averaging 1320 hours of use per year. Here we are seeking funding to purchase the next generation of spinning disk confocal microscope to increase the capacity that the facility can offer and maximise throughput. The microscope we have chosen will also bring automated, AI-driven imaging and analysis capabilities to the facility. The facility’s users routinely image whole brain and organ slices from a variety of model organisms, as well as intact organoids, to increase our basic understanding of neurodevelopment, neurophysiology and organogenesis. For imaging of large areas our point-scanning confocal microscopes force a trade-off between resolution and total time required for acquisition. We wish to purchase the Nikon Crest-Optics X-LightV3/DeepSIM spinning disk microscope which produces the largest images, with the flattest illumination field currently available, utilising the full 25mm field of view in the optical light path. The incorporation of two Photometrics Kinetix sCMOS cameras will allow simultaneous dual-colour imaging, enhancing the speed of acquisition of fixed tissue images.  This new imaging technology will reduce our imaging time by 10-fold compared to our point scanning confocals with no sacrifice in resolution. Furthermore, this new technology will allow UoE to develop spatial transcriptomics pipelines, integrating projectome mapping (the network of neural connections) with transcriptome profiling (gene expression data), advanced techniques currently restricted to a few labs exclusively in the United States. Spinning disk microscopes are well regarded for minimising phototoxicity while imaging live cells and tissues. The combination of spinning disk technologies with the fast imaging rates of Kinetix cameras (>1000fps) will allow the simultaneous recording of integrated, highly dynamic events, such as image based voltage sensing and calcium signalling, within cells and tissues. The Crest-Optics DeepSIM super-resolution modality, combined with the spinning disk, will create an imaging system capable of imaging across scales, contextualising subcellular events within a tissue. The hardware control is fully integrated within NIS-Elements. Combining the hardware with the AI image processing features available within Nikons-Elements, and the JOBS graphical programming package, will allow automatic imaging protocols to be developed to identify and image rare or event-driven instances with fixed or live tissue. The system will be housed in a dedicated imaging facility managed by two full time members of staff who are imaging and image analysis specialists. The facility business plan allows us to keep all of our instruments under fully inclusive service and repair packages, minimising uncertainty of downtime of any of our systems. The IMPACT facility is part of the Edinburgh Bio-Imaging Consortium which brings together over 30 bio-imaging and analysis experts from across Edinburgh and Heriot-Watt Universities. We use this critical mass of expertise to provide an all-encompassing support network for our users, sharing equipment and resources, and collectively negotiating best-value for procurement.

UKRI Gateway to Research · FY 2025 · 2025-07

Global cloud-resolving models (GCRMs) underpin the study of many important application areas, such as wind energy and extreme weather hazards, related to the physics and dynamics of weather regimes. Modern supercomputers have the computational power to run GCRMs at one-kilometre scales. In this project, we investigate whether reducing grid spacings to one kilometre or less increases the model accuracy and whether the dynamics and the physics at those spacings, for example, concerning convection, cloud microphysics, or turbulence schemes, should be reviewed. This will advance knowledge about the physics and dynamics of weather regimes at scales of 100 to 1000 m and use novel model implementations to analyse wind patterns over complex terrains and at sea. A key challenge is understanding whether it is possible to reduce the grid spacings of GCRMs by leveraging the next generation of computing infrastructure, including exascale and GPU accelerators. The project seeds an ongoing collaboration between three international partners with complementary expertise. The Department of Wind and Energy Systems at the Technical University of Denmark (DTU) is a leader in atmospheric flow observations and meteorological models that can be used to study the impact of the atmosphere on wind energy systems. The Physical Oceanography Department at the Center for Scientific Research and Higher Education at Ensenada (CICESE) operates regional meteorological forecasts and weather and climatological databases for the Mexican Northwest and has led the Consortium for Research in the Gulf of Mexico since 2015. EPCC at the University of Edinburgh are leaders in the application of High-Performance Computing to research, co-developed the Met Office NERC Cloud Model (MONC), and host the UK's national supercomputer, ARCHER2. Improving the resolution of GCRMs will play a critical role in advancing our understanding of clouds and their interactions with the broader climate system, by permitting the resolution of convection and turbulence in the "gray zone" of atmospheric scales between 100 and 1000 m at Global and Regional scales for periods longer than one season. This will improve forecasting and modelling, studying extreme weather events, and aid in the planning of renewable energy systems.

UKRI Gateway to Research · FY 2025 · 2025-07

Bacteria rule the Earth. Over billions of years, these fantastically diverse and abundant organisms have established and maintained Earth's nutrient cycles, infused the oceans and atmosphere with oxygen, and formed, shaped and weathered rocks and minerals. Even today, they are the true foundation of our biosphere. But fossil bacteria are tiny, ambiguous, often poorly preserved, and therefore challenging to study. Now, for the first time, new technology is beginning to reveal their secrets. Advances by our research group and others provide tantalising new data about the structure and make-up of fossil bacteria, but the interpretation of these data is vexed by uncertainties about the processes of decay and mineralization (fossilization) that ultimately produced them. Our work to decode fossil bacteria will reduce these uncertainties and thereby advance major debates about the origins of important groups and their relationships to key events in Earth's history. Firstly, by showing experimentally how selected, ecologically important groups of bacteria are broken down and replaced or encrusted by minerals in the laboratory (making "artificial fossils"), we will discover new clues that palaeontologists can use to recognise these groups in the fossil record. This in turn will advance the study of ancient ecosystems and environments. Secondly, we will combine these experimental insights with sophisticated chemical microscopy to update our understanding of some of the best preserved fossil bacteria known to science, which occur in the Rhynie chert in Scotland. These world-famous rocks open a spectacular window onto life on land 400 million years ago, including some of the oldest well preserved land plants and diverse bacteria. We will conduct a high-resolution census of Rhynie chert bacterial diversity, testing hypotheses about microbial life during the greening of the land. For example, many of the filament-shaped bacteria have been identified as (oxygen-making) cyanobacteria, but our knowledge of modern spring communities predicts that diverse non-cyanobacterial filaments should also be present, playing different, potentially important, ecological roles. Thirdly, we will address an apparent billion-year discrepancy in the fossil record of nitrogen-fixing cyanobacteria, helping to constrain the history of Earth's atmosphere. The rise of oxygen forced some cyanobacteria to evolve specialised cells (heterocysts) for nitrogen fixation in aerobic conditions. The timing of this consequential innovation is poorly understood: heterocystous cyanobacteria also produce distinctive resting cells (akinetes), but the oldest purported fossil akinetes pre-date by ~1 billion years the oldest known heterocysts (e.g., those in the Rhynie chert). Because these different cell types differ fundamentally, they should decay and preserve quite differently. We will test whether differences in preservation potential can solve the puzzle of the missing heterocysts in the fossil record: perhaps they are older than they seem. This project will create fundamental new knowledge of ancient bacteria and new methods for studying them, opening a new frontier in palaeontology. Our findings will engage bio- and geo-scientists through multidisciplinary journals and conferences, and will have implications for the discovery and analysis of Earth's oldest fossils, the search for life on Mars, and the co-evolution of life and Earth through deep time. We will engage non-specialists in the UK and beyond through public talks, school visits and resources, work with Rockwatch (the club and magazine for young UK rockhounds), a new "fossil bacteria" article for Wikipedia, one of the world's most viewed websites (which provides pageview data), and press releases.

UKRI Gateway to Research · FY 2025 · 2025-07

In the past 15 years, there have been significant technological advances in the instrumentation available for imaging cells and tissues without the addition of fluorescent labels. This is important as the addition of these labels can change the behaviour of the system itself, obfuscating any biological results. These advances have led to increases in the speed of imaging, the level of detail obtained and the information content of these so-called “label-free” technologies. Advances in image interpretation driven by artificial intelligence enable the analysis of ever larger and more detailed experiments. Hence, we are entering an era where label-free imaging can be used to analyse detailed changes in complex, live cell and tissue models. Of all the label-free techniques, Coherent Raman Scattering (CRS) microscopy has been identified as a “Method to Watch” (Nature Methods, 2022) for imaging cells and tissues. CRS confocal microscopy captures quantitative data about the strength of discrete Raman vibrations corresponding to individual bonds across each pixel in an image. The morphology of a cell can be readily visualised in 3D by tuning to specific carbon-hydrogen (C-H) bond vibrations that reflect the environment of these bonds in proteins or lipids. In its hyperspectral mode, CRS microscopy generates pseudo-Raman spectra for every pixel within the desired field of view with unrivalled spatial resolution, providing an information-rich approach to microscopy and a powerful tool for chemical analysis. The proposed commercial multiphoton/CRS confocal microscope will allow world-leading scientists in the Institute for Regeneration and Repair (IRR) at the University of Edinburgh to map the distribution of different molecular species, probe changes in molecular composition, directly image small molecules and apply a range of image analysis techniques in live cell and tissue models - label-free - using an accessible, reliable platform for the first time. Currently there is only one equivalent commercial multiphoton/CRS microscope in the UK, which is not in a biology-led facility. The University of Edinburgh is world-leading in its use of CRS microscopy in the biosciences and has received funding for its development from UKRI, industry and charities. The IRR core Imaging Facility has recently expanded its label free imaging resources and with our CRS expertise we are exceptionally well placed to build on and expand our bioscience capabilities through investment in a commercial multiphoton/CRS microscope. The microscope will be a new tool that will enable us to address fundamental bioscience questions including: identifying hallmarks of cellular ageing; determining mechanisms of lineage commitment in tissue engineering; sensing intracellular conditions during tissue repair; and predicting causality in bioimage analysis. The acquisition of a commercial multiphoton/CRS microscope will immediately benefit researchers across the University of Edinburgh and Scotland, but will also be advertised to UK users through the UK Technology Specialists Network and imaging forums such as the Royal Microscopical Society, and the Society of Photo-Optical Instrumentation Engineers. By embedding the microscope in the IRR Imaging Facility, we will ensure that its continuity of operation and ease of access to all users is underpinned by trained professional staff. The impact of the research it enables will include an enhanced understanding of fundamental process which govern life; a reduction of animal use in research (through the development of more meaningful human cell and tissue models); and the development of cell and tissue models of healthy, aged and diseased tissue for biomedical research.