University College London

UKRI Gateway to Research · FY 2025 · 2025-11

Context: Microbes are found everywhere, from the soil and the seas to the our food and bodies, performing fundamental roles in nature and human health. Advances in molecular and synthetic biology have demonstrated that engineered microbes have the potential to be a crucial component in future sustainable technologies from manufacturing to medicine. Despite their potential, using single types of engineered microbes (monocultures) has limitations. There is only so far that we can push cells before they succumb to evolutionary and other pressures, and our engineering efforts are lost. It has long been proposed that division-of-labour, in which a complex and resource intensive task is divided into more manageable chunks and distributed to specialised workers, can provide the solution to monoculture limitations. Challenge: Placing different microbial strains into the same environment leads to competition for space and resources, resulting in extinction of some strains, and failure of the larger process. Strains in wild microbiomes are able to co-exist by interacting with one another, demonstrating both combative and cooperative behaviours. While some simple examples of engineered microbial communities have been created by manipulating such interactions, the ad hoc approach taken to-date has provided very little foundation on which to build communities for real applications. Aims and objectives: This project aims to move beyond ad hoc demonstrations of simple communities, to develop a deeper understanding of how the various genetic tools at our disposal can be harnessed to efficiently and effectively produce robust microbial communities for biotechnological applications. To achieve this we will: Develop a platform for the design, implementation, and assessment of synthetic microbial communities; uniquely bringing together threads from mathematical and biological disciplines, and bridging them with computational and robotic capabilities. Through rapid prototyping and iteration of new microbial communities, explore factors that contribute to community stability and resilience; propagating this knowledge back into the community design process. Demonstrate the application of this platform and knowledge to a complex bio-process; producing high-value biochemicals from sugars in a four-strain community Potential applications and benefits: The direct outputs of this project will provide immediate impact in fields such as ecology and microbiology, where existing top-down approaches to understanding community behaviours have focused on analysing wild microbiomes or assembling novel communities of wild strains. This project will deliver the tools for a systematic bottom-up approach, enabling falsification and refinement of theory that has been difficult to achieve in the past. Our proof-of-principle community bio-process, and the systematic engineering framework, will also provide short term benefits to those interested in advancing bio-industrial applications. In the longer term, we aim to use this as a foundation for understanding how engineered microbes interact with existing microbiomes; a fundamental challenge and important responsibility as we push engineering biology applications into medical and environmental settings. Given this breadth of possibility, this project will provide impact from academia to industry and broader society. Relevance to BBSRC priorities: This project aligns with the BBSRC’s goals of advancing bioscience for sustainable agriculture, health, and advanced manufacturing; providing a foundation for immediate and sustained improvement across all of these areas. It also supports UKRI’s mission to drive innovation and economic growth through cutting-edge science and technology. By addressing current limitations in microbial engineering, this project contributes to the UK government’s focus on engineering biology as a critical technology.

UKRI Gateway to Research · FY 2025 · 2025-11

Context: In the rapidly evolving world of healthcare, it is crucial for the UK to maintain its edge in developing and delivering cutting-edge medicines swiftly and sustainably. AstraZeneca, the UK’s largest biopharmaceutical company, in partnership with UCL Biochemical Engineering, aims to tackle these challenges through a strategic Prosperity Partnership. Building on a decade-long collaboration, this partnership leverages a rich history of joint achievements, including significant contributions to bioprocessing research and the training of key industry and academic leaders. The Prosperity Partnership initiative focuses on next-generation therapies - such as multi-specific antibodies, antibody-drug conjugates (ADCs), and protein nanoparticles - that offer tremendous potential but come with complex manufacturing demands.   Challenge: The production of these advanced therapies faces significant hurdles, including low yields, intricate processes, and stringent quality requirements. Traditional methods are often insufficient for these novel treatments, demanding innovative approaches to streamline their development and manufacturing processes.   Aims and Objectives: The aim of our partnership is to transform the way we make medicines. By combining advanced AI technology and digital tools, we plan to make the development of biopharmaceuticals quicker, more cost-effective, and environmentally friendly. Our goal is to ensure that new treatments reach patients faster and to significantly reduce the carbon footprint of manufacturing these medicines, thereby setting new global standards for how drugs are produced.   The project is structured into four interconnected work packages (WPs): WP1: Focuses on using AI and engineering biology for rapid screening and optimization of cell lines, directly influencing the quality and quantity of biopharmaceutical drug production. WP2: Aims to improve the ease of manufacture of novel biopharmaceuticals, ensuring scalability and robustness in production. WP3: Utilizes digital twins and Process Analytical Technologies (PAT) to simulate and optimize manufacturing processes, enhancing decision-making and efficiency. WP4: Evaluates the economic and environmental impacts of new manufacturing technologies, promoting sustainable practices.   Potential Applications and Benefits: Our partnership is poised to make a significant impact on healthcare and the environment. Quicker Access to New Treatments: By streamlining the development and manufacturing processes with AI and digital technologies, new therapies can be produced faster. This means patients might have access to the latest treatments sooner than ever before, potentially saving more lives and improving the quality of life for those with chronic conditions. Environmental Impact: Our efforts to achieve net-zero carbon emissions in biopharmaceutical medicines manufacture represent a significant step towards sustainable healthcare. Reducing the environmental footprint of medicine production not only aligns with global sustainability goals but also sets a new standard for the pharmaceutical industry worldwide. Economic Growth and Innovation: This project not only supports the growth of the UK’s biotechnology sector by enhancing its competitiveness globally but also fosters innovation through the development of cutting-edge manufacturing technologies. As these technologies are adopted more widely, they will drive down costs and improve efficiency, benefiting the entire sector. Educational and Professional Development: The partnership ensures that the next generation of scientists and engineers are trained in the latest digital and bioprocessing techniques. This prepares them for high-impact careers in the biopharmaceutical industry, contributing to the overall skill level and expertise within the UK and beyond.   Our partnership demonstrates a commitment not only to transforming biopharmaceutical development but also to doing so in a way that is responsible and forward-thinking, ensuring benefits for society at large and the environment.

UKRI Gateway to Research · FY 2025 · 2025-11

This project aims to bridge ecology and bioinformatics to provide a novel way of understanding and predicting the stability of cellular ecosystems within tissues—a crucial step toward understanding tissue health and resilience, with broad applications to agriculture, animal and human health. Cells interact dynamically with neighbors and their environment, both locally (e.g., T cells responding to neighbors) and over longer distances (e.g., chemokines guiding immune cells to infection sites). While much research has focused on understanding how individual cells communicate, how these interactions collectively shape tissue behaviour remains poorly understood. This knowledge is crucial for understanding of both healthy tissue function and how tissues break down or transform in disease. Advances in technologies like spatial omics and high-resolution imaging have revolutionized our ability to study these relationships.  Spatial omics allows us to map gene expression patterns within tissues, essentially giving us a snapshot of what cells are doing and how they interact across regions. Despite this progress, we still lack robust methods to analyze these interactions in a way that can predict how tissues behave under different conditions. This proposal addresses this gap by creating a new cross-disciplinary approach merging bioinformatics with principles from ecology—a field that has long studied complex systems and interactions. The parallels with ecological research are clear. Ecological communities consist of individuals of many different species that interact with different strengths over restricted distances. Combined with restricted movement, this leads to emergent complex spatial patterns that can tell us much about the underlying biological processes that we do not view directly. This knowledge is crucial to predicting how these species and communities will respond to perturbations like climate shifts. By applying ecological approaches at the cellular level, we will create a toolkit for investigating tissue organization and breakdown under normal and altered conditions to answer questions such as how tissues maintain their structure, how they respond to injuries, and what happens when they undergo changes due to diseases like cancer. Our research has two goals. First, we will develop methods to define cellular communities that function as “mini-ecosystems” within tissues. Ecological communities are often formed of sub-regions typified by certain subsets of species, such as vegetation types across a landscape. These spatially localised communities emerge due to differential interactions between different species within the broader region. Parallel definitions can be used to define cellular communities. By profiling gene activity across tissues, we can map the areas where cell interactions drive emergent behaviour. To identify these areas, we will adapt ecological methods used to map distinct communities in natural environments. Testing these methods in simulated and real datasets will allow us to refine our approach, ensuring it accurately captures cellular communities in various tissue types and organisms. Secondly, we will link these patterns to biological processes and create predictive models that assess how cellular ecosystems respond to disturbances. In natural ecosystems, certain communities can withstand significant changes (like forests recovering after fires), while others may shift to a new state entirely. Similarly, some cellular ecosystems within tissues may be resilient, while others may be prone to breakdown under stressors like aging, injury or disease. By simulating disturbances in these communities—such as removing specific cell types or altering cell-to-cell interactions—we aim to predict the points at which tissues might transition to new, stable states.  

UKRI Gateway to Research · FY 2025 · 2025-11

The BIO-SHARP project proposes a unique and transformative approach to invasive medical procedures. It addresses the precision challenges associated with needle insertion in the Seldinger techniques, used for catheter placement in blood vessels and hollow organs, particularly the Central Venous Catheter. These procedures, crucial to modern healthcare, often come with substantial risks, including high complication rates that can lead to severe health consequences. With over 27 million of these procedures performed globally each year, and complication rates reaching as high as 20%, the need for enhanced safety and efficiency in these methods is urgent.   The project will pioneer the BIO-SHARP robotic system, an integrated needle platform with the potential to revolutionise invasive medical procedures. It will incorporate bio-impedance analysis onto a tiny needle empowered by integrated circuits. The needle can detect and differentiate tissue layers using bio-impedance data. Through machine learning, the robotic system converts the bio-impedance into tactile sensations, relaying critical subcutaneous information to the clinician's fingertip, allowing for intuitive and precise needle insertion to the intended location.   With this new tactile experience guided by bio-impedance, our state-of-the-art engineered needle system aims to significantly reduce clinicians' cognitive and physical demands, allowing for a higher level of focus and fewer equipment adjustments posed by ultrasound guidance. As a result, the risk of complications such as needle misplacement or puncture could be greatly reduced, leading to safer and more efficient medical procedures.   The BIO-SHARP project will collaborate closely with industry partner Haply Robotics, which specialises in haptic technology, to develop the visioned haptic-assistive system. Additionally, we will conduct workshops and pre-clinical studies focusing on the Central Venous Catheter procedure with University College London Hospitals to validate and refine the system's effectiveness.   The objectives of the project are: Develop the integrated needle sensory platform. Develop the haptic-assistive BIO-SHARP system for needle insertion. Facilitate workshops and pre-clinical studies in collaboration with hospital partners to promote and validate the effectiveness of the BIO-SHARP system.   The BIO-SHARP system can potentially have applications spanning various medical procedures, from cardiac or arterial catheterisation to percutaneous tube insertions and spinal or epidural anaesthesia. Its versatility promises to enhance patient care, improve healthcare cost efficiency, and boost the overall efficacy of medical practice. The BIO-SHARP system could also greatly simplify the training process and pave the way for advanced techniques to become more accessible to healthcare providers.   The BIO-SHARP project aligns with the rapid growth of surgical robotics, propelled by healthcare challenges such as an ageing population and the ongoing impact of COVID-19. The project aims to redefine standards in Seldinger-related invasive medical procedures and influence the trajectory of its medical training, potentially leading to far-reaching benefits for the healthcare system, improving patient outcomes and working conditions of medical professionals, and marking a significant step forward in the integration of technology in healthcare.

UKRI Gateway to Research · FY 2025 · 2025-11

In animal cells, cell division is driven by a series of precisely orchestrated shape changes that couple the segregation of genetic material to cytokinesis, the physical separation of the two daughter cells. Understanding the molecular and mechanical processes underlying cytokinesis has fascinated biologists for decades. Much attention has focused on the mechanics of the actomyosin cortex, a thin submembranous network of F-actin and actin-binding proteins. Myosin motors generate contractile tension in the cortical network, and their accumulation at the equator is thought to drive the cleavage of the mother cell during cytokinesis. Equatorial cortex accumulation is associated with at least partial alignment of actin and myosin into a ring-like structure, an organisation thought to promote force generation. However, the extent of this alignment and its importance for cytokinetic mechanics remain unclear. Furthermore, changes in surface tension are coupled to changes in cell volume, which increases at mitotic entry and decreases progressively as the furrow forms and ingresses. Volume changes likely also change cytoplasmic mechanical properties, but how this affects cytokinetic mechanics has not been investigated. Finally, plasma membrane reorganisation and changes in intercellular adhesions also likely contribute to cytokinetic mechanics. Taken together, increasing evidence points to the importance of considering the mechanics of the cell as a whole to understand the dynamic shape changes underlying cytokinesis. This proposal aims to leverage recent developments in advanced imaging, cellular mechanics, and image analysis to interrogate the interplay of the mechanical properties of various cellular components during cytokinesis, and to ask how together, they drive robust shape changes underlying daughter cell separation.   We will focus on four aims: -Aim 1 will explore the use of morphometric analysis to analyse mitotic shape changes. In this approach, cell shape is quantified through a large number of features, which are then mapped into a low-dimensional “morphospace” using dimensionality reduction. We will then integrate morphospace analysis with quantifications of the distributions of F-actin and myosin to generate morphomolecular trajectories of cytokinesis. -Aim 2 will characterise spatiotemporal changes in the mechanical properties of cellular components during cytokinesis. We will map mechanical changes onto morphomolecular trajectories to provide hypotheses for the role of cortical tension, membrane tension, and cytoplasmic mechanics in shape change. -Aim 3 will investigate the molecular and biophysical mechanisms controlling cytokinetic shape changes. We will determine how perturbation of key cytoskeletal proteins and the contractile ring affect morphomolecular trajectories. We will then perturb volume regulation and the control of membrane tension to determine their respective contributions to mitotic morphogenesis. -Aim 4 will examine the mechanics of cell division in epithelia, where cells must maintain barrier function while changing shape. We will quantify forces arising at intercellular junctions throughout cytokinesis using DNA hairpins as sensors. We will incorporate these resistive forces into our computational simulations and interrogate potential mechanoresponsive feedbacks during cytokinesis in an epithelial context. Finally, we will probe the biophysical limits of the trade-off between achieving robust cytokinesis and preserving barrier integrity by modulating the strength of intercellular adhesions with DNA nanotechnology. Together, our research will generate a holistic understanding of the multiple mechanical processes that participate in mitotic morphogenesis and will dissect the molecular mechanisms controlling force generation and shape change dynamics required for robust and successful cell division.

UKRI Gateway to Research · FY 2025 · 2025-11

Ageing is the main risk factor for neurodegenerative disorders, including Alzheimer’s disease and associated dementias. Yet, we still do not understand the molecular mechanisms involved. It is now increasingly evident that ageing is highly heterogeneous and even cell-type specific, requiring a reappraisal of its role as a contributing factor for neurodegeneration at the cell-type level. Loss of proteostasis is a hallmark of the ageing brain and it affects pathways involved in the clearance of pathological proteins linked to dementia, such as tau. This project will first define/identify loss of proteostasis at the cell-type level using spatial biology methods. By performing a targeted approach to compare vulnerable and resilient cell-types, I will unravel novel pathways of cellular resilience and vulnerability to neuropathology and investigate the consequences of cell-type specific proteostasis blockage on neuronal function. Upon completion, this project will provide a mechanistic understanding of proteostasis loss during ageing which will be used as a framework for the development of novel biomarkers and therapeutic development.

UKRI Gateway to Research · FY 2025 · 2025-11

Context: Depression, anxiety, and attention deficit hyperactivity disorder (ADHD) are common mental health conditions that occur from childhood to adolescence and have major impacts on young people’s well-being, development, and educational and economic outcomes. The Challenge Discovering effective interventions to mitigate the adverse consequences of these conditions requires reliable causal evidence. However, limitations of existing evidence include: Limited evidence from RCTs: There are relatively few randomised trials of young people and those that do have small sample sizes, limited follow-up, and do not record critical outcomes of interest to patients and the public. Observational studies can be unreliable because controlling all pre-existing differences between individuals (confounding), reverse causation, and differentiating familial factors is challenging.  Conventional Mendelian randomization using population-based summary data can be unreliable because it does not differentiate societal and familial factors from causation. This project aims to leverage large-scale genetic data to better understand the causes of these conditions and inform the development of effective interventions to improve young people's mental health. It will use a novel Mendelian randomization phenome-wide association study (MR-PheWAS) within family Mendelian randomization and drug-target Mendelian randomization analyses. Aims and approaches Aim: To provide reliable causal evidence of mechanisms that affect young people’s mental health. This project will develop reliable evidence of causal mechanisms and potential intervention targets and strategies to improve outcomes for young people with 1) depression and anxiety and 2) ADHD using three approaches: Approach 1: Discover when the earliest symptoms of depression, anxiety and ADHD emerge and evaluate the direction of causation for risk factors. Approach 2: Investigate how familial factors affect young people’s mental health. Approach 3: Evaluate existing drugs and discover new targets for improving young people’s mental health. Secondary objectives Apply, adapt and validate existing pipelines and methods for depression, anxiety and ADHD. Disseminate our research findings. Develop and deliver advanced methods workshops and short courses. Impacts Advance scientific knowledge by providing new evidence about modifiable environmental risk factors and potential effective interventions.  Apply rigorous methods by developing and applying three novel molecular genetic techniques, including MR-PheWAS, within family and drug-target Mendelian randomization. Provide new evidence from some of the largest family-based studies in the world, such as the Norwegian Mother, Father and Child Cohort Study (MoBa), Millennium Cohort Study (MCS), and Born in Bradford (BiB).  Include diverse populations by analysing ethnically diverse samples from BiB and MCS, improving the diversity of genetic epidemiological studies.  Direct and indirect benefits Build collaborations via interdisciplinary collaborations with researchers with expertise in clinical psychiatry (Argyris), psychology (Pingault), quantitative sociology (Bann), population genetics (Badini), genetic epidemiology (Hemani), social epidemiology (Howe), econometrics (Sanderson), and internationally, clinical psychology (Havdahl). Training and capacity building, particularly for early career researchers and graduate students, via short courses and tutorial papers. Dissemination of findings and best practices via pre-registration, code review, reproducible code publication, and guidelines for genetic epidemiological analyses. How will this project improve human health and clinical practice? This project will improve our understanding of the causes of depression, anxiety, and ADHD in young people using longitudinal and genetic data. The findings will inform the development of future interventions, ultimately improving outcomes for young people, their families, and society.

UKRI Gateway to Research · FY 2025 · 2025-11

Treatment-Resistant Depression (TRD) is a severe form of depression that does not respond adequately to at least two different antidepressants. It causes significant emotional distress, making everyday tasks such as work, relationships, and basic self-care incredibly challenging. Despite its profound impact, little is known about why some individuals respond to antidepressants while others experience TRD, or about the long-term effects of TRD on overall health. My research explores the relationship between TRD and metabolic health. Metabolic health involves how your body manages blood sugar, cholesterol, blood pressure, and fat. Poor metabolic health can lead to type 2 diabetes and heart disease. There are several reasons why TRD and metabolic ill-health could be linked. Firstly, living with poor metabolic health can be emotionally challenging, which may worsen depression and hinder treatment. Secondly, untreated depression can make it harder to lead a healthy lifestyle, resulting in poor metabolic health. Third, biological factors, such as inflammation or hormone imbalances, could influence both metabolic health and the brain’s response to antidepressants. Fourthly, difficulties in everyday life – financial struggles, unstable housing, lack of education or job opportunities, limited access to support, distressing environments – may make it harder to keep physically and mentally healthy. I aim to understand the relationship between TRD and metabolic health, whilst considering the wider context of people’s lives. I aim to identify changes we can make, both clinically and socially, that will help make depression easier to treat in people with poor metabolic health — and will in turn help people with depression maintain better metabolic health.

UKRI Gateway to Research · FY 2025 · 2025-11

The Late Veneer (LV) is one of those fundamental paradigms in the formation and subsequent evolution of Earth that endured since the 1970s. In simple terms, the LV refers to a small amount of currently unknown material added to the Earth after its initial formation and the effective closure of the core. The evidence for the occurrence of a LV rests almost entirely on the interpretation of HSE (highly siderophile element) abundances in the mantle. Based on expected partitioning behaviours, these elements should have been almost entirely sequestered into the metallic core during planetary differentiation, leaving the Earth’s mantle devoid of them. In reality though, the abundance of HSEs in the upper mantle is at least three-orders of magnitude higher than expected. This requires a source of HSEs to the Earth after core-closure. Hence, the “Late Veneer” hypothesis. However, the partitioning data which underpins the LV model is from experiments at much lower pressures (P) and temperatures (T) than those extant at core-formation. Moreover, available data show such a large effect of P and T on this partitioning that the core may not be so effective at sweeping up these metal-loving elements as previously thought. It is, therefore, entirely possible that a LV is not needed at all to explain observations. Resolving this important issue would have far-reaching consequences for Earth accretion and force a re-set in our understanding of planetary formation. The aim of this project is to use state of the art ab initio calculations to retrieve partition coefficients of all the HSEs (which include Os, Ir, Ru, Rh, Pt, Pd, Au, and Re) between the core and mantle under the extreme P/T conditions appropriate for planetary differentiation. We do this because such partition coefficients are difficult or impossible to achieve experimentally. The ultimate vision is to use these new results to assess whether there is a need for a Late Veneer at all, or whether something more evolutionarily complex is required. This new work will provide robust constraints on the processes by which the Earth accreted and evolved to its current state. We foresee its future use in the interpretation of HSE inventories for other sampled silicate bodies (i.e. Mars, Moon, Vesta).

UKRI Gateway to Research · FY 2025 · 2025-11

STFC Science faces systemic diversity issues (e.g. post-16 physics is <30% girls; low-income background and Black students are highly underrepresented in STFC science). The UK also faces chronic shortages of science teachers (1 in 7 schools does not have a physics teacher) leading to ongoing damaging skill shortfalls. To address these problems, Orbyts creates partnerships between researchers and schools. The researcher acts as a relatable role model, having an ‘immensely humanising impact’ on perceptions of science and scientists, while supporting the school students to lead their own original research projects.  Orbyts’ structure of regular interventions, relatable role models and active ownership of research is proving to be transformative: Evaluation shows that Orbyts is significantly increasing science capital and empowering school students to engage in STEM pathways in the long term. Schools that host Orbyts projects report increases in STEM uptake at A-level and beyond following a project (we encourage the panel to see: www.orbyts.org/impact). School teachers report that without Orbyts’s partnership model they would not have time, subject expertise and/or the confidence to deliver these types of projects. Consequently, Orbyts caters to schools that cannot undertake similar programmes (e.g. the excellent IRIS). Orbyts is an initiative that takes the burden of running impactful student research projects off of time-starved, overworked teachers, while providing them with unique Continuous Professional Development. Over the past 9 years of careful growth, we have facilitated 150+ different, bespoke research-with-schools projects, that have enabled more than 250 school students to author scientific publications on STFC science; the majority of these students are from groups under-represented in science (https://www.orbyts.org/researchprojects-publications). The introduction of a National Coordinator in May 2024 has given Orbyts the capacity to expand further ensuring high quality projects and, through extensive reflection and evaluation, we have begun to build a unique body of best-practice on the (co)creation and delivery of impactful research-with-schools partnerships. Orbyts Aims to leverage long-term researcher-school partnerships and the excitement of STFC research to: A1) Empower school students with tools, sense-of-agency and confidence to pursue STEM subjects and related careers pathways. A2) Increase STEM uptake by students from historically-excluded groups A3) Improve scientific literacy and provide high-level STEM training for school students A4) Provide relatable STEM role models for school students, and dispel harmful stereotypes regarding who is suitable for scientific careers A5) Provide our overstretched teachers with unique, exciting CPD opportunities that inspire their continued love of their subject and raise teacher retention A6) Provide researchers with transferable management, pedagogy and communication skills, laying the firmest foundations for the next generation of UK STEM lecturers and leaders Objectives: Our previous Spark Award enabled us to launch a now-thriving Midlands hub. This Nucleus award builds on this to grow and connect a national Orbyts community, to: Deliver a National programme of Orbyts research-with-schools partnerships  Produce 3 peer-reviewed studies that explore and identify the impact of Orbyts-like research-with-school partnerships on Key Stage 3/4 school students from underrepresented groups, teachers and the researchers that deliver the projects. This will share the nature of the impact with the wider (inter)national community. Collate, document, characterise and disseminate the best practice  Each hub partners with local schools to develop deep community networks locally, while sharing best practice across the national network of hubs at the Universities of Leicester, Warwick, Northumbria, Kent, Sheffield and University College London.

UKRI Gateway to Research · FY 2025 · 2025-10

Artificial Intelligence (AI) systems will fundamentally transform how scientific evidence is accessed by policy makers. As governments increasingly adopt AI-powered systems for evidence identification, conventional pathways for connecting science to policy will be disrupted. This represents a major shift in the science-policy interface that has been essentially unchanged for decades. While research on general AI ethics and societal impacts has grown rapidly, a significant empirical gap exists in understanding precisely how AI will alter which scientific evidence informs governmental decisions. This project addresses a fundamental question at the intersection of AI, metascience, and evidence-based policymaking: to what extent and in what ways does AI-sourced evidence differ from human sourced evidence in policy contexts? As governments transition toward AI-mediated evidence ecosystems, we need empirical analysis of whether AI systems recommend substantially different scientific evidence than current human-driven processes. These differences could either amplify existing limitations in evidence selection or potentially correct blind spots in human-curated evidence. I seek to conduct a large-scale, data-driven comparison between human-selected and AI-selected scientific evidence in policymaking. This analysis will create counterfactual comparisons between evidence historically cited in policy documents and what would have been recommended by AI for identical policy questions. I will quantify systematic differences in evidence selection patterns across the United Nations Sustainable Development Goals (SDGs) and governmental bodies, while measuring shifts in researcher networks between those currently cited in policy versus those identified by AI systems.   To achieve these objectives, I will leverage a novel AI recommendation system developed during my doctoral research. In partnership with Elsevier, who will provide comprehensive metadata and access to their vast scientific database, I will apply this tool retrospectively to study differences between actual policy citations and AI-recommended research across the SDGs. I will do this in two stages, first by identifying a case study of systematic literature reviews conducted for a policy issue, and comparing what these reviews identified against what AI would identify, then based on lessons learnt, scale the analysis using Overton policy citation data to over 2 million policy documents worldwide spanning the SDGs. Finally, I will perform a meta-analysis examining patterns of AI vs human cited research. This research will deliver a quantitative analysis that measures the gap between AI-recommended and human-selected evidence across different SDGs, identifying which domains are most susceptible to disruption as AI systems are adopted. Second, a comprehensive dataset mapping the "invisible college" of researchers who produce highly policy-relevant work, but aren't currently recognised in policy citations. This will enable universities to identify and support overlooked talent that current human-driven processes miss.

UKRI Gateway to Research · FY 2025 · 2025-10

Circadian rhythms time organismal behaviours across the day-night cycle. A key function of the circadian clock is to regulate the onset and offset of sleep, a behaviour essential for neurological function. Perturbed sleep-wake patterns are frequently observed in patients with neurodevelopmental disorders, creating significant burdens for their carers. More broadly, sleep defects in healthy adults, in aged humans, and in patients with neurodegenerative disorders, represent an increasing socio-economic problem. Hence, there is an urgent need to understand the molecular mechanisms that promote circadian rhythmicity, and thus healthy sleep patterns. Circadian rhythms are highly conserved across metazoan species, and genetically tractable model organisms have yielded important insights into their molecular underpinnings. Here, we utilise the fruit fly, Drosophila, as such a model, and present robust preliminary data that defines a protein called Rbf as a novel and unexpected regulator of circadian rhythms and sleep timing. Rbf plays well-known roles in suppressing DNA synthesis and G1-S phase progression during the cell cycle, and a conserved orthologue is present in humans (RBL2). We reveal a previously unappreciated role for Rbf in post-mitotic circadian (clock) neurons that no longer undergo cell division. We show that loss of Rbf in clock neurons strongly disrupts circadian locomotor rhythms in adult flies. Additional cell-specific manipulations suggest a direct effect of Rbf on the molecular clock, and that Rbf is required in multiple clock neuron subtypes for rhythmic locomotor behaviour. To define cellular roles of Rbf in post-mitotic neurons, we utilised transcriptomic approaches. These revealed that loss of Rbf dramatically upregulates expression of DNA synthesis genes in mature neurons. These transcriptional changes correlate with non-uniform increases in DNA content and nuclear size across adult fly neurons, suggesting that specific neuronal subtypes in the fly brain are particularly vulnerable to genome replication (polyploidy) following loss of Rbf. Based on our preliminary data and prior studies, we therefore hypothesise that loss of Rbf similarly induces genome replication in a non-uniform manner across clock neurons. This, in turn, causes the molecular clock in distinct clock neuron subsets to run at different paces, leading to network desynchronisation, circadian arrhythmicity, and altered sleep timing. Through our Work Packages we will robustly test this hypothesis, thus defining the molecular mechanisms through which Rbf promotes circadian rhythmicity. Our recent study shows that human patients with mutations in the Rbf orthologue RBL2 exhibit disrupted sleep patterns, demonstrating an important and conserved role for this protein across evolution, with direct relevance to human health. More broadly, cell-cycle dysregulation is increasingly recognised as an important feature of the ageing brain and in neurodegenerative disorders. By examining how cell cycle disruption in mature clock neurons impacts circadian rhythmicity, our work therefore has relevance to understanding sleep disruption in these medically and socio-economically important conditions. Finally, by showing that Rbf continues to suppress cell cycle gene expression after cell cycle exit, our work reveals a new mechanism by which neurons maintain their post-mitotic identity across the lifespan – a process fundamental to neurological function.

UKRI Gateway to Research · FY 2025 · 2025-10

Prions are infectious proteins which cause fatal brain diseases in humans and other mammals, the most well-known of which is Creutzfeldt-Jakob Disease (CJD). Prions have unusual biological properties, including the ability to reproduce themselves (self-propagate) without the need for DNA or similar molecules. These behaviours were always considered unique to a single protein called the prion protein, but recent events have challenged these assumptions. The amyloid-beta protein causes two human diseases: Alzheimer’s disease, a dementia that causes memory and thinking difficulties; and cerebral amyloid angiopathy (CAA), which causes haemorrhagic (bleeding) strokes. In the last decade, doctors have identified patients who have developed Alzheimer’s disease and CAA at an unusually young age. All these patients had medical procedures many years earlier, often in childhood, and most procedures involved “cadaveric” (i.e. provided by people after death) material. These young patients developed Alzheimer’s disease or CAA because abnormal amyloid-beta protein was transferred to them via this cadaveric material, resulting in disease decades later. The very same medical procedures are known to have caused iatrogenic (medically acquired) CJD after transfer of abnormal prion protein by this route. The existence of these iatrogenic forms of Alzheimer’s disease and CAA, together with three decades of animal model data, strongly suggests that amyloid-beta can sometimes act as a prion. Although these iatrogenic forms are likely to be rare, the disease mechanisms causing them might be important for “sporadic” (without apparent cause) forms of these conditions, which are much more common, particularly at older ages. If this is the case, prion biology provides a unique framework for understanding how and why amyloid-beta causes disease, and could help identify new strategies for prevention and treatment. The programme of research described in this Fellowship has two aims. The first is to better understand iatrogenic forms of Alzheimer’s disease and CAA by working with patients with these conditions, as well as people at risk of developing them (i.e. those with known exposure to cadaveric material, specifically cadaveric human growth hormone). The second aim is to expand the prion experimental toolkit for amyloid-beta, with the eventual goal of developing a range of biological assays equivalent to that for the prion protein. This experimental toolkit would have several applications and could be used to test mechanistic hypotheses, including whether Alzheimer’s disease and CAA can be caused by different versions (strains) of amyloid-beta. It could also be used to build biological plausibility and address important outstanding public health questions about other procedures which might transmit amyloid-beta. This is particularly relevant to the question of whether amyloid-beta diseases can be transmitted by blood transfusions, as there are data from a large Scandinavian study (over 1 million participants) that suggest this might be possible. The MRC Prion Unit at UCL recently hosted a workshop attended by a range of stakeholders, including senior representatives from the Department of Health and Social Care, UK Health Security Agency and NHS Blood and Transplant, to discuss this possibility; our discussions highlighted the need to develop robust experimental methods to explore this question. The projects in this fellowship will take the next steps in developing this experimental toolkit for amyloid-beta, and will apply new biological assays to explore the questions of amyloid-beta strains and blood-borne amyloid-beta transmission.

UKRI Gateway to Research · FY 2025 · 2025-09

Context NHS hospitals hold extraordinary health data that could accelerate medical breakthroughs and improve patient care. Yet accessing this data remains painfully slow. Researchers wait months for approvals, then work within secure systems that constrain collaboration. More importantly, patients and the public have little visibility of the health information held within these enclaves. Synthetic data offers a solution: artificial datasets that mimic real patient data without containing actual patient information. But Current tools fall short. They are either opaque "black boxes" that lack the transparency NHS information governance teams need to assess risks, or oversimplified systems that cannot handle the complexity of electronic health records. NHS information governance teams need transparent systems. NHS analysts need tools designed for real-world healthcare data. Both need practical pathways that work within existing NHS structures. Our Solution: Progressive Data Layers Like a Russian nesting doll (Matryoshka), we propose progressive layers of health data using OMOP, the interoperable format being adopted across UK health systems. Public Synthetic Data: Created transparently from summary statistics using differential privacy. Straightforward approval because generation methods are fully visible. Valuable for understanding data structures, testing methods, and education with no access controls needed. Private Synthetic Data: Generated using complex algorithms that learn from individual patient records whilst creating entirely artificial data. Less transparent generation but maintains privacy budgets. Provides high-fidelity data for serious research with lighter access controls than secure systems. Real Data: Actual patient information within existing secure controlled environments, reserved for final research stages after full approvals. Aims and Objectives In partnership with patients and public, NHS information governance and data teams, privacy experts from the Alan Turing Institute and UCL's leading Trusted Research Environment group, we will provide comprehensive technical tools and governance frameworks for progressive synthetic health data sharing. We will: Extend our proven technology (SQLSynthGen) to handle the complexities of OMOP, the standard for hospital electronic health records Create transparent governance models enabling NHS information governance teams to make safe, rapid decisions about synthetic data releases Establish systematic evaluation methods determining what research can be accomplished with different synthetic data types Build sustainable training programmes for NHS data analysts and information governance teams Demonstrate real-world applications connecting London NHS partner hospitals to our regional secure data environment Develop reusable templates and workflows for adoption across UK health systems Potential Applications and Benefits We focus on implementation, not theory. Excellent synthetic data tools exist, but practical pathways aligned to NHS Information Governance and the skills of NHS data management teams are missing. We have both a proven technical implementation at University College London Hospital working with >1 million health records, and proven governance frameworks already being rolled out NHS hospitals across North Central London. Our vision is that every NHS healthcare provider maintains a publicly accessible synthetic version of its data assets. This enables open conversations between academic and industry researchers, NHS staff, and patients. Research teams can coordinate partnerships and rapidly design analyses externally before accessing secure environments. Results from private data can be checked and validated transparently. Excellent synthetic data tools already exist, but excellent synthetic data pathways aligned to NHS Information Governance approvals, and the tools and skills of NHS data management teams are missing. We have proven we can solve this locally, and we now seek support to extend this nationally.

UKRI Gateway to Research · FY 2025 · 2025-09

Following high-profile incidents like the Lakanal House and Grenfell Tower fires, which exposed how seemingly routine building maintenance enact violence towards local populations, this project addresses the urgent need to retrofit high-rise buildings in ways that are both environmentally sustainable and socially responsible. With over 16% of EU emissions coming from the domestic sector, and high-rises being particularly "complex-to-decarbonise," retrofitting these structures is crucial. However, the process presents significant socio-technical challenges, including fire hazards and moisture problems that compromise structural integrity and resident health, as evidenced by the Grenfell Inquiry. These challenges are not merely technical but deeply social and cultural, embedded in the lived realities of inhabitants and the institutional frameworks that govern retrofitting. Using ethnographic and visual research methods, I will engage residents, architects, and policymakers to create retrofit solutions that prioritise safety, well-being, and cultural integrity. The project combines an ethnographic study of a retrofitted high-rise in Lewisham, South London, with participant observation inside the Lewisham Council’s Climate Resilience team, which developed the borough’s Housing Retrofit Strategy. This dual approach captures both residents' lived experiences and the institutional framework guiding retrofitting, offering a comprehensive view of its impacts and governance. Lewisham’s diverse community, grassroots activism, and ageing high-rise housing stock provide a rich context to explore how environmental goals intersect with residents' experiences and the borough's unique social and historical dynamics. The unique anthropological contribution lies in developing a new theoretical framework that reframes retrofitting as a practice of "infrastructural care," foregrounding the interplay between material changes to buildings and the social, cultural, and emotional dimensions of residents' lives. The research aims to create this framework by integrating architectural transformations with residents’ social relations, cultural practices, and embodied experiences, bridging technical demands with the lived realities of residents. Research questions: How do residents of ageing high-rises in London engage with the retrofitting process, and how do their values and practices intersect with or challenge the sustainability priorities of local authorities? In what ways can retrofitting practices be reimagined as a form of infrastructural care that aligns technical imperatives with residents’ cultural, social, and emotional needs? How can retrofitting strategies address the dual goals of environmental sustainability and social equity, fostering meaningful collaboration between residents and local councils? By identifying the challenges that climate change poses for ageing buildings and examining the infrastructural care required to address them, this study opens a new interdisciplinary approach to retrofitting. The project’s objectives are: Empirical: Develop an evidence-based understanding of residents’ experiences of retrofitting, exploring their views, values, and critiques. Compare tenants' concerns with those of the council to examine how climate resilience strategies are implemented and perceived. Conceptual: Expand and integrate anthropological theories on the home, infrastructure, and climate change, and develop architectural theory through an anthropologically grounded approach to retrofitting as infrastructural care. Engagement: Intervene in public and policy debates through co-creative methods such as ethnographic documentary and participatory workshops, fostering collaboration between residents and local authorities. Practical: Transform retrofitting practices by developing a socio-technical framework for a sustainable building industry. This co-creative approach will produce actionable frameworks for local councils and housing authorities, enhancing policy, cross-sector collaboration, and a sustainable building industry. The adaptable model developed here will provide strategies for retrofitting that address climate resilience and social equity in urban housing across the UK and beyond.

UKRI Gateway to Research · FY 2025 · 2025-09

AI/ML Training (DiRAC) This proposal presents a structured training programme designed to equip UK researchers with essential skills in the artificial intelligence (AI) and machine learning (ML), technologies that are rapidly transforming scientific research. While AI/ML offer powerful tools for solving complex problems and driving innovation, many researchers lack the technical expertise required for effective and responsible use. This skills gap poses risks to research integrity and scientific discovery. To address this and drive the wider adoption of AI capabilities across research fields,  the proposed initiative will deliver a modular, self-paced training programme through the established DiRAC Training Academy online platform. The curriculum will focus on core machine learning techniques, including deep learning and generative models, and will be inclusive and freely accessible to the UK research community. Interactive Jupyter Notebooks using real scientific datasets will provide hands-on experience and promote awareness of high-quality, well-governed  data sources. The programme will also feature a series of expert-led “deep-dive” sessions. These ~2-hour recorded lectures will explore real-world applications of AI/ML within specific scientific domains, reinforcing theoretical knowledge through practical demonstrations of successful research codes and methods. To further support participants, the final phase of the programme will include live virtual drop-in sessions and interdisciplinary discussion forums. These will offer one-on-one expert guidance, facilitate cross-disciplinary collaboration, and encourage co-creation of innovative research approaches. Overall, this initiative will build a confident, skilled, and diverse research community, accelerate the integration of AI as a standard scientific tool, and foster long-term innovation across UKRI domains.

UKRI Gateway to Research · FY 2025 · 2025-09

This innovative PhD project seeks to address the long-standing issue of strayed records in the UK National Archives (TNA) by leveraging crowdsourcing. Despite existing preventative measures, thousands of records—including books, maps, and documents—become separated from their original collections. These records often become separated due to mishandling or oversight. While archivists manage and catalogue records, researchers bring expert knowledge of specific subjects, which can help identify and reunite these separated records.  The current reunification process is slow and resource-intensive, necessitating an innovative approach to prevent the backlog from growing. By engaging both archivists and archive users, the project seeks to develop a novel method to reunite strays in archival collections. The objective is to develop a sustainable method, minimising staff time and resources while keeping a good balance of accuracy and engagement with the public of the archive. The research addresses key questions: how can crowdsourcing effectively assist in reuniting strayed archival records? What other methodologies can enhance the efficiency and accuracy of this process? how can a sustainable, scalable model be developed for long-term impact? Maintaining record integrity is crucial for historical research. While crowdsourcing has been used in the Galleries, Libraries, Archives, and Museums (GLAM) sector, this project is unique in applying it to a management problem. Given TNA’s successful crowdsourced transcription projects, this is an opportune moment to harness public expertise for reunification. This project will be of interest to many within heritage conservation and management, GLAM, digital heritage and digital humanities communities. This project will be of interest to diverse communities researching our past. This project is in line with TNA’s research vision and priorities. It cuts across two areas of expertise identified in the Research Vision: 1) Digital, Data and Emerging Technologies 2) Impact, Culture and Engagement. This project will use emerging technologies, i.e. the crowdsourcing platforms available due to the Internet, to solve one of the longstanding management challenges of archives. This project has the potential to establish TNA as a world-leading research organisation in crowdsourcing to solve management challenges. This project will use emerging technologies to empower archive users. One of the greatest benefits of crowdsourcing is users worldwide can interact with the collections without necessarily visiting TNA. This project will actively engage with volunteers to ensure that they feel valued and their contributions are recognised. This project will make a real-world impact. It will contribute to long-term goals of TNA. This research will offer an innovative, scalable solution to a persistent archival challenge faced by archives worldwide while strengthening connections between archives and their user communities. A new subfield of research within crowdsourcing in heritage will emerge from this project with involving users in solving longstanding management challenges in GLAM. The impact of this project will remain even after the project finishes.  It will also impact thousands of archives in the UK and beyond. Long-term real-world impact will be achieved through expediting the process of reunification, saving time and resources and helping in preserving memories for the users of archives.

UKRI Gateway to Research · FY 2025 · 2025-09

We are developing a new generation of quantum sensors to enable earlier diagnosis through ultra-sensitive tests to detect lower biomarker levels, and improved access through lower costs and portable instruments. Despite major breakthroughs, quantum sensors still face barriers before their full potential for healthcare applications can be realised by deploying them in in hospitals, clinics, or low-resource settings. Current experiments can require long measurement times, carefully controlled lab environments, and expert manual analysis. These constraints threaten to prevent widespread use in real diagnostics. Even where the fundamental sensitivity is high, turning that sensitivity into real clinical performance (particularly in messy, variable environments) is an ongoing challenge. There is a pressing need to integrate artificial intelligence (AI) to make quantum sensing technologies faster, smarter, and more adaptable to complex real-world conditions. This project brings together researchers from various disciplines and UK institutions to use AI to address these challenges for quantum sensors in five key areas: Making quantum sensors smarter and faster: We will use AI to reduce the number of measurements needed, by identifying and focusing only on the most informative data points. This will reduce scan time and improve performance in noisy and unpredictable conditions. Boosting in vitro diagnostics: We are developing spin-enhanced lateral flow tests that use nitrogen-vacancy centres in nanodiamonds, and have demonstrated a 100,000-fold improvement in fundamental sensitivity over the current standard method. To translate this into real-world assays, we will use AI for automated image analysis that can improve sensitivity and distinguish between specific and nonspecific binding. Improving epilepsy detection: We will use AI to automatically detect epileptic brain activity from traditional magnetoencephalography and quantum (optically pumped) magnetoencephalography. Discovering better materials for diagnostics: Using AI to explore and optimise new fluorescent nanomaterials for use in blood-based cancer monitoring. This will allow faster, more accurate detection of circulating tumour DNA in blood samples. Driving the wider adoption of AI: Hosting training courses and bootcamps for researchers interested in learning how to apply AI techniques for their own research.

UKRI Gateway to Research · FY 2025 · 2025-09

The UK’s rich collection of longitudinal population and cross-sectional studies form the backbone of empirical research in the social, economic and behavioural sciences, as well as in epidemiology and health research, providing the basis for evidence-based policy advice. However, the demands of the research community, encapsulated by the FAIR principles, for more timely and higher quality granular metadata is a significant challenge. This project focuses on addressing these barriers to uplift existing metadata resources and lay the basis for future data to be truly FAIR and usable by AI technologies. The project has four main objectives: Extraction of metadata from social survey questionnaires to increase both volume and interoperability of these data sources to enhance discovery. Knowledge enhancement through concept extraction and classification into standard vocabularies which are supported and used by the social science, health and statistical community. Development of pipeline to integrate outputs into a FAIR enabled semantic web (RDF) store / knowledge graph to store and make accessible the metadata generated for use by AI technologies, development of quality assurance tooling. Utilisation of these enhancement metadata to develop new approaches to tackle privacy and disclosure challenges to enable improved decision based and for this to be scaled through automation. These build on the pilot/development work supported by the UKRI/ESRC Future Data Services over the last 12 months. The structured data resources created can be leveraged to create new ways of utilising rich Social Science resources for discovery, harmonisation and comparability, removing previous barriers such as language and non-interoperable formats, providing: Standards-based provenance metadata (how the data was collected) Annotation to standard controlled vocabularies Automation of privacy and access rights How the proposal meets AI for Science Objectives The project would develop metadata extraction and metadata enhancement technologies which would  be robust across the known potential data sources and make a significant contribution to the volume and availability of such resources for AI. The development of an RDF metadata  store which was curated with community vocabularies would enable new approaches to social science data, not viable with current levels of fragmentation, over and above those already developed in this work. How the proposal meets DSIT Objectives Policy makers are  increasingly demanding that analysis is provided in a more timely and comprehensible manner. At the same time, data is coming from more diverse sources, and needs to be combined across disciplines and organisations. We believe that enhanced metadata generated from these work streams, which is structurally and semantically coherent and made available to analysts alongside the data, is a critical part of assuring them that they are looking at data that is trustworthy and can be combined in a meaningful way. The workstream on disclosure is one example of how these metadata can be further utilised to move very challenging manual tasks to (initially semi-) automated decision based processes.  

UKRI Gateway to Research · FY 2025 · 2025-09

Many societal and technological challenges can be addressed by discovering and designing novel materials. Materials are employed everywhere, from prosthetics and drug delivery in health care to electronic devices to nuclear waste storage, carbon capture, and catalysis to reduce our impact on the environment. The development and design of new advanced materials can help to tackle challenges in key areas such as catalysis of complex chemical reactions, how charge and heat travel through devices, energy generation and storage (batteries or fuel cells) and recovering waste heat (thermoelectric materials). Computational materials modelling methods are now accurate enough to lead this work, but these simulations use a large fraction of high-performance computers worldwide.   Machine Learning (ML) methods provide efficient ways to model complex phenomena and can be used to accelerate materials simulations, post-processing, and analysis of the results. This enables scalable, cost-effective and responsive computer modelling across materials science, which supports real-time decision making, modelling of larger and more realistic materials (for example, to include complex defects present in real materials that either provides the property of interest or is the source of its degradation), and a greater understanding of the “uncertainties” in the simulation for example by understanding the effects of the computer approximations.   The next generation of supercomputers exploit GPUs, as CPU-only machines are less energy efficient. Many ML methods also run efficiently on GPUs, ultimately making materials simulations using them even more energy- and cost-efficient. We have already shown that this kind of ML-based approach is effective at analysing Small Angle X-ray Scattering (SAXS) data (https://doi.org/10.1021/jacsau.4c00368), complementing computationally intensive Monte Carlo simulations, for predicting materials properties that are usually computed using expensive electronic structure methods (https://doi.org/10.1038/s41524-024-01486-1) and even for generating candidate structures for global optimisation approaches (https://doi.org/10.1038/s41467-024-54639-7).  In this work, we will embed ML methods within the popular CASTEP materials modelling software, which is developed in the UK and has a worldwide userbase, including many industrial R & D teams. We will investigate a variety of different ML methods from deep neural networks to generative AI to determine the most suitable approaches to address both fundamental science questions and specific high-priority applications, as identified by the UK materials research community (e.g. the UKRI high-priority use-cases) and piloted by us in a recent joint CCP9 and CCP-NC feasibility study of predicting the electron density in a material using Machine Learning. We will also investigate where else AI can be exploited to help the materials and molecular modelling community and either immediately act on these, plan future research packages, and/or disseminate our findings to our community. For example, we will (a) link at least one our materials software codes to the AI code already developed by the Alan Turing Institute to enable an easier route for exploiting AI routines; (b) survey where AI is already being exploited within and outside our community; and (c) develop how to capture atomic structural coordinates of predicted clusters reported in early publications as images of ball and stick models and make the structural data available to the community via our web-accessible database for cluster structures. The main aim is to embed AI and ML methods in materials research tools in a fair and inclusive way, to create user-friendly, high-impact software which brings the advantages of AI to materials research in a responsible and reliable way.

UKRI Gateway to Research · FY 2025 · 2025-09

The Challenge The UK's journey to net zero depends on understanding how people use energy in their homes, but accessing the data needed for this research is restricted. Smart meter data from households provides crucial insights for developing better energy policies, reducing fuel poverty, and improving building performance. However, privacy regulations and data protection rules create major barriers to accessing this information, slowing down vital research that could accelerate our transition to clean energy. Our Solution We will solve this data access problem using cutting-edge artificial intelligence. Our project will train advanced AI models called Generative Pretrained Transformers (GPTs) on the UK’s world-leading SERL Observatory dataset - a unique collection of smart meter data from 13,000 representative GB households, combined with detailed information about their buildings, occupants, and energy use patterns collected over five years. The AI models will learn the complex patterns in real energy data and generate completely synthetic datasets that look like real household energy data but contain no actual personal information. This synthetic data will have all the statistical properties researchers need while being completely privacy-preserving. This project directly advances EPSRC AI for Science objectives: developing AI capabilities across research fields to accelerate scientific discovery; increasing access to well-governed, high-quality datasets for AI; building interdisciplinary collaborations between AI and energy researchers; and embedding AI as a research tool in a fair and inclusive way. What We Will Deliver Over six months, we will create "Synthetic-SERL", the first dual-fuel synthetic smart meter dataset with long temporal sequences. This will include half-hourly gas and electricity data for 13,000 virtual households across full calendar years, each with contextual information about building and occupant characteristics and weather. We will rigorously test this synthetic data to ensure it provides genuine research utility while passing strict privacy audits. The entire dataset, along with the training code and tools to integrate the data into workflows, will be published under open licences, making it freely available to researchers worldwide. Impact and Applications This research will deliver targeted benefits across three key sectors. Academia will benefit from removal of barriers to accessing high-quality data, enabling and accelerating R&D that was previously restricted by privacy regulations. Industry will gain access to data for testing business cases and grid planning that rely on high-resolution energy data, supporting investment decisions in clean energy technologies. Government will have data to evaluate distributional impacts of net zero policies like time-of-use tariffs, ensuring a fair transition for all households. By democratising access to high-quality energy data, we will unlock research that was previously restricted, accelerate innovation in the energy sector, and create new partnerships between AI researchers and energy experts. This project establishes the foundation for future research that will push the boundaries of AI for energy decarbonisation.  

UKRI Gateway to Research · FY 2025 · 2025-09

Amyotrophic lateral sclerosis (ALS) is a fatal and incurable neurodegenerative disease characterised by the loss of upper and lower motor neurons. The majority of ALS cases are sporadic and characterised by cytoplasmic mis-localisation of TDP-43, an RNA-binding protein that is predominantly nuclear under normal physiological conditions. The loss of nuclear TDP-43 results in the aberrant inclusion of intronic regions called “cryptic exons” (CEs) in mature transcripts. CEs typically lead to the loss of the protein they are encoding and the disease relevance of these events has been established, with therapies targeting them currently in clinical trials. Recently, we and others have described CE events that can be translated, therefore leading to the formation of proteins containing a novel “cryptic peptide”. Whilst the potential of such proteins as biomarkers is actively being pursued, whether they also drive motor neuron degeneration, and the molecular mechanisms by which this might occur, remain critical knowledge gaps. Through RNA-seq data analysis from post-mortem tissues, we detected the expression of a CE in the alanyl-tRNA synthetase 1 (AARS1) mRNA specifically in the spinal cord and motor cortex from ALS patients with TDP-43 pathology. Intriguingly, the AARS1 CE is inserted in-frame, yielding a variant of the AARS1 protein that contains a cryptic peptide (AARS1Cryptic). Importantly, missense mutations in AARS1 have been linked to Charcot-Marie-Tooth disease type 2N, a hereditary peripheral neuropathy where motor axons are affected. Since AARS1 CE inclusion occurs specifically in regions where motor neurons degenerate in ALS and single amino acid changes in AARS1 can perturb motor neuron integrity, AARS1Cryptic, where a novel peptide is inserted and predicted to alter a crucial region of the protein, represents an intriguing candidate driver of ALS pathology. In this proposal we present an experimental strategy to examine how expression of AARS1Cryptic contributes to ALS. Utilising human iPSC-derived lower motor neurons (i3LMN) and the fruit fly, Drosophila melanogaster, we have generated unique in vitro and in vivo models expressing the AARS1 CE-containing protein (AARS1Cryptic). Our preliminary data indicate that AARS1Cryptic exhibits reduced protein translation/stability and a failure to localise to motor neuron axons. Furthermore, neuronal overexpression of cryptic (but not wild-type) AARS1 reduces night sleep in Drosophila, revealing that AARS1Cryptic disrupts a fundamental neurological process (sleep), recognised to be an important risk factor for neurodegenerative diseases. Using these models, we will test whether AARS1Cryptic expression can disrupt neurological function and modify neurodegenerative phenotypes caused by TDP-43 loss-of-function. Our research strategy promises to rigorously test whether AARS1Cryptic can drive neurodegeneration and/or modify TDP-43 pathology; define how this CE affects the enzymatic activity of AARS1 and its protein interactome; and gain further insights into the molecular mechanisms contributing to motor neuron loss downstream of TDP-43 mis-localisation. Our programme will thus advance the fundamental understanding of ALS biology and widen the range of potential therapeutic targets to treat this devastating disease.

UKRI Gateway to Research · FY 2025 · 2025-09

We propose to develop efficient, low cost and reliable mid-infrared (3 to 15 µm) interband cascade lasers (ICLs) on Si substrates for many applications such as chemical sensing, greenhouse gas detection, environmental monitoring, detection of contaminations in food commodities, and industrial process control. These ICLs are desirable and enabling components for mid-infrared (IR) sensor systems for Mid-IR Si photonics. The research will build on our newly proposed innovative quantum well (QW) region containing strained InAsP layers for ICLs and on extensive experience and achievements in ICL structures and devices, as well as on the molecular beam epitaxy (MBE) growth and device fabrication of Si-based quantum dot lasers. ICLs using the new QW active region with InAsP layers have been demonstrated by our University of Oklahoma team with significantly improved device performance in the long wavelength region (10-14 µm) compared to the commonly used regular W-QW active region, suggesting enhanced optical gain. Hence, we will apply this innovative QW active region to ICLs in a wide mid-IR spectrum (3-10 µm) to significantly improve their device performance. Additionally, we will use advanced waveguide configuration with hybrid cladding layers (with both highly doped semiconductor plasmon layer and superlattice or quaternary AlGaAsSb layer) and grow ICLs on Si substrates, which have much reduced cost and notably higher thermal conductivity compared to GaSb or InAs substrates. As such, ICLs will have improved optical confinement and enhanced thermal dissipation, resulting in further reduced power consumption and higher output power, as well as lowered cost. Here it should be mentioned that University College London (UCL) MBE group is one of a few MBE facilities with As, P and Sb capacity within one MBE chamber worldwide, and only one in the UK. The approach and tasks involve design and modelling of ICLs, molecular beam epitaxial growth of device structures, material characterization, and device fabrication and testing. These ICLs will significantly benefit many useful applications, especially where high power is required or mid-IR systems must be operated with batteries and energy cost/availability is a concern, including space applications with strict constraints on size and electric power. The availability of high-performance ICLs will greatly enhance the capabilities of mid-IR laser instruments and their applications in many areas. As these lasers will be fabricated on Si substrates, it will offer a route of monolithically integrate Mid-IR laser sources for Si electronics for a wide arrange of applications from sensor, detection of pipe leaks and explosives, food safety, medical diagnostics, and industrial process control, to space applications. This project explores the quantum engineering of novel semiconductor heterostructures to demonstrate and develop such much-needed semiconductor lasers. It is well lined up with UK National Semiconductor Strategy and two EPSRC challenge themes, Manufacturing the Future and Digital Economy. The objectives of the project also include to advance the understanding and knowledge of the new heterostructures and their behaviours in semiconductor lasers.  The project offers graduate and undergraduate students unique opportunities to pursue education, training and research in multidisciplinary topics. This project not only advances the understanding and knowledge of semiconductor sciences, but also generates new knowledge in the design of quantum-engineered structures and broadens their applications.

UKRI Gateway to Research · FY 2025 · 2025-09

Context and Challenge The special educational needs and disabilities (SEND) system in England is in crisis. Evidence-based solutions are required to ensure that children’s SEND needs are identified early so that they can receive timely support and improve their educational outcomes (Curran,2020). Early SEND identification and support can also provide much-needed, long-term cost savings. This is because children can receive the support they need from an early age before issues become well-entrenched and more difficult to address as they get older (Carneiro et al.,2024). Universal screening measures, which assess children’s cognition, language, socio-emotional, and physical development, play an important role in the early identification of SEND (Miles et al.,2018). This is because effective universal screening measures can efficiently identify which children may need additional follow-up assessments, interventions, and support (APPG,2023). While statutory measures, such as the Early Years Foundation Stage Profile (EYFS-P), are implemented with all young children in England, there is little evidence evaluating how effective some of these universal screening measures are for identifying children’s future SEND needs (Atkinson et al.,2022;Snowling et al.,2011; Wood et al.,2024). Aims and Objectives Our study proposes to conduct the first national-scale evaluation of two universal screening measures in early childhood for identifying later SEND. The two measures are: The EYFS-P, which is a statutory observational assessment completed by teachers with all 4-5-year-olds in England at the end of Year Reception. An adapted version of the Vineland Adaptive Behaviour Scale, which is a questionnaire completed by parents when children are 3 years old in ‘Understanding Society’, a nationally representative dataset. We will address three core objectives. First, our study will evaluate the extent to which children’s performance on the EYFS-P at age 4-5 can predict their later SEND needs up till age 15-16, including for children from different ethnic, socio-economic status, and English as an Additional Language groups. Second, our study will estimate, of the children who were aged 0-4 years during the Covid-19 pandemic, what proportion are at a heightened risk of later SEND needs based on their EYFS-P performance. Our study will also estimate how much additional educational funding is required to support these children, which will inform future DfE spending reviews. Third, our study will evaluate the extent to which even earlier indicators of child development at age 3 (i.e., the adapted Vineland Adaptive Behaviour Scale) can predict later SEND needs. Our study will also make comparisons between the two universal screening measures, which will inform evidence-based guidelines for educational policymakers and practitioners. We will answer these questions by conducting quantitative analyses of administrative (the National Pupil Database) and large, nationally representative survey (Understanding Society) datasets. Potential Applications and Benefits Our new evidence will be used to advise educational policymakers and practitioners across England about how children’s SEND needs can be identified from an early age. Recommendations will also be made regarding the funding required to support the educational needs of children who experienced much of their early childhood during the Covid-19 pandemic. Our interdisciplinary research team have world-leading expertise in early years education, SEND, and secondary data analysis. We will work with our extensive networks in the DfE and Ofsted, as well as other non-government organisations and early years practitioners, to ensure our findings are used to address the SEND crisis and make a tangible difference for young children.  

UKRI Gateway to Research · FY 2025 · 2025-09

Spina bifida (SB) is a common birth defect affecting annually over 250,000 pregnancies worldwide. Folic acid supplements reduce the SB rate but, even when added to the food supply (‘fortification’), folic acid prevents fewer than half of SB cases. With the growing availability of fetal surgery, more families are opting to have a live-born child with SB rather than abortion. However, the education and eventual independence of children with SB are limited not only by their lower body disability, but also by learning difficulties. These result from the Chiari II brain disorder which includes hydrocephalus and higher brain defects that affect learning. Fetal surgery for SB reduces hydrocephalus risk, but does not protect against the higher brain disorders. This research proposal aims to determine how the higher brain defects of Chiari II arise in embryos with SB. We are in a unique position to do this research, having recently developed the first mouse model in which spontaneous (genetic) SB leads to the higher brain defects of Chiari II. Our mouse has a genetically normal head, and so the presence of Chiari II brain defects demonstrates they arise secondary to the SB, potentially via leakage of cerebrospinal fluid from the open spinal cord. Our first results show that the higher brain defects result from faulty neuron production in the early embryo with SB. In Aim 1 of the proposal, we will examine the early stages of neuron formation in the brains of embryonic mice with SB, compared with normal littermates. We will identify whether a global disorder of neuron formation is present, or whether only certain neuron/glial types are affected. This information will reveal the fundamental basis of the Chiari II brain defects. In Aim 2, we will explore the underlying molecular basis of this disorder, using the contemporary method of ‘spatial transcriptomics’. This combines RNA sequencing (to identify which genes are expressed) with maintenance of the spatial structure of the tissue, in this case the brain. Hence, we aim to discover which genes are differently expressed in the various neuron-forming regions in brains of embryos with or without SB. This analysis will identify which of many molecular brain signalling pathways may be affected by the presence of SB, and which could be targets for novel therapies. In Aim 3, we will use mouse embryos that are grown outside the mother’s body. Using this system, we can close the SB at a much earlier stage than in human fetal surgery. This will test whether closure prevents the higher Chiari II brain disorder. Then we will trial drug treatments, at this same early stage, aimed at rectifying the disorders of neuron formation, and therefore preventing the brain anomalies. Overall, this project has the potential to significantly advance our understanding of the brain disorders in SB that affect so many children’s lives. Our prevention studies may form a proof-of-principle for subsequent studies in larger animals and, eventually, may benefit human pregnancies with SB, especially by improving the outcome for fetal surgery.