143174 research outputs found
Sort by
Assessing the UK Labour Government’s Tax Policies in Its First Year in Office
No full textThe file attached to this record is the author's final peer reviewed version. The Publisher's final version can be found by following the DOI link.The article examines the tax policies the UK Labour government pursued in its first year in office in the context of its economic and fiscal inheritance. It presents the promises of the 2024 Labour Party Manifesto and the government’s main tax policies to June 2025. The article considers several possible interpretations of the government’s tax policies including the desire both to raise revenues and to establish fiscal credibility and assesses whether tax measures have been used to address social policy goals. The article concludes that, while cautious and even conservative, Labour’s approach to tax and tax policy making did not appear to be based on a cohesive plan, was not accompanied by a coherent or compelling narrative, and failed to seize the opportunity for a bolder strategy
Heterogeneity in public attitudes and preferences for the deployment of aquifer thermal energy storage
No full textAquifer thermal energy storage (ATES) can contribute to heating and cooling decarbonization by utilizing the thermal capacity of natural aquifers. Securing acceptance and support for deploying ATES at scale requires acknowledging public perceptions and designing systems compatible with public preferences. Here we characterize attitudinal stances and preferences for the deployment of ATES in public buildings in the UK. Using data from a social survey and a discrete choice experiment, we find substantial heterogeneity in public attitudes and support for ATES installations. Latent class analysis identifies four distinct stances, ranging from cautiously negative to enthusiastically supportive. Estimating mixed multinomial and hybrid choice models, we find strong preferences for quicker deployment of ATES infrastructure, with greater CO2 emissions-reduction capacity, that can be accessed by private households. Results point to the need for tailored communication strategies and preference-compatible design for achieving socially desirable sustainable energy transitions
Deep learning for wireless communications
No full textDeep learning (DL) has shown its overwehlming privilege in many areas, including computer vision, robotics, and natural language processing, where it is normally difficult to find aconcrete mathematical model for feature representation. Inspired by many application exampls of DL, especially the success of AlphaGo in 2015, we initialized DL for communication in 2016 to completely or partially address the following issues in wireless communications.Open Acces
Real-world evidence on RSV vaccine uptake, effectiveness, and safety in older adults: a systematic review and meta-analysis
No full textBackground
Vaccines to prevent respiratory syncytial virus (RSV)–associated lower respiratory tract disease in older adults have become available in recent years. We investigated RSV vaccine uptake, effectiveness, and safety signals in older adults reported in post-licensure real-world studies.
Methods
For this systematic review and meta-analysis, we conducted 11 monthly searches (between November 5, 2024, and November 10, 2025) in Ovid Medline, Embase, and Global Health databases. Meta-analyses, using random-effects modelling, were performed for uptake, effectiveness, and safety signals. PROSPERO registration: CRD42025643585.
Findings
A total of 3900 studies were identified, of which 36 were included, published between December 22, 2023, and October 28, 2025, and covering over 121.8 million individuals across the United States, United Kingdom, Italy, Australia, Czech Republic, Switzerland, France, Canada, and Israel. In the United States, RSV vaccine uptake among adults aged ≥60 years during the 2023/24 RSV season was 18.0% (95% confidence interval (CI): 12.2–25.7; ten studies), varying by clinical and socio-demographic subgroups. Among adults aged ≥60 years, pooled estimates of vaccine effectiveness were 75.3% (95% CI: 73.7–76.9; three studies) against any laboratory-confirmed RSV-positive infection, 76.4% (95% CI: 74.2–78.5; four studies) against RSV-related emergency department or urgent care visits, 74.8% (95% CI: 66.8–82.9; six studies) against RSV-related hospital admissions, and 79.8% (95% CI: 68.1–91.5; four studies) against severe RSV-associated disease. Following vaccination, Guillain-Barré syndrome (GBS) was reported in two studies with between 5.2 and 6.5 cases per one million doses for RSVPreF3+AS01 (Arexvy, GSK) vaccines and between 9.0 and 18.2 cases per one million doses for RSVpreF (Abrysvo, Pfizer) vaccines.
Interpretation
RSV vaccine uptake in older adults was low globally with substantial disparities between sociodemographic and clinical subgroups. Our study showed a favourable safety profile and high effectiveness of the RSV vaccines, highlighting the value of wide implementation of these vaccines.
Funding
There was no funding for the study
ReverseGWAS identifies combined phenotypes associated with a genotype in GWA studies
No full textMotivation
Traditional genome-wide association studies (GWAS) aim to uncover the genetic variants associated with a single phenotype of interest (typically a disease), and to elucidate its genotypic architecture. However, many of today’s GWAS simultaneously measure multiple related phenotypes, leading to the possibility of pursuing the reverse aim of elucidating the “phenotypic architecture” of a single genetic variant. In other words, we may ask what combination of measured phenotypes is associated with a given genotypic variant. ReverseGWAS is an algorithmic platform for answering such questions in the context of large-scale multi-phenotype GWAS.
Results
We demonstrate the effectiveness of ReverseGWAS on simulated data, showing its ability to identify logical combinations of phenotypes with a reasonable amount of noise. We then apply it to a selection of combined phenotypes from the UK Biobank, obtaining 719 candidate associations using autoimmune diseases and 205 using common ICD10 codes. We find that the majority of these associations (546/719 and 111/205, respectively) successfully replicate in an independent cohort, FinnGen.
Availability
The source code of ReverseGWAS is freely available to non-commercial users as an installable R package at https://github.com/Leonardini/rgwas.
Supplementary information
Supplementary data are available at Bioinformatics online
A reproducible integrating sphere method for standardised diffuse light characterisation of photoelectrochemical and photovoltaic devices
No full textPreprint versio
Growing the UK’s AI assurance market in defence and security
No full textExecutive Summary
This CETaS Briefing Paper provides an evidence-based analysis of the UK’s AI assurance market for Defence and National Security (D&S). A thriving AI assurance sector could enable AI adoption and become a key driver of UK economic growth. If organisations are confident that AI harms can be mitigated, this will help the UK Government achieve its aim of “fast, wide and safe” adoption of AI1 and prevent AI capabilities from failing at the implementation stage. Effective assurance processes allow for the rapid integration of AI into existing business structures and processes, which can contribute to economic growth across multiple sectors. However, a range of factors currently limit both the supply of and demand for AI assurance services. 1 HM Government, AI Opportunities Action Plan (Department for Science, Innovation and Technology: January 2025).
This Briefing Paper describes the current state of AI assurance in national security, defence and policing organisations, highlighting its strengths, challenges and possible mitigations. Drawing on this, the paper identifies lessons from D&S to support the growth of a robust AI assurance market for other sectors and advance AI innovation across the UK economy. Key findings from this study are as follows:
• D&S is a diverse sector in AI assurance maturity, as it includes both early adopters and organisations at the start of their AI assurance journeys. This is due to a range of factors that vary across the sector, including: level of AI adoption; technical skills; infrastructure and testing capabilities; risk appetite; preference for inhouse or external offerings; and level of engagement with external providers of AI assurance.
• Demand for AI assurance in D&S is driven by a desire to secure strategic and operational advantage from effective AI, the risk of high-consequence errors, policy requirements, and a need to assess AI providers’ claims. Demand for third-party assurance is driven by skills shortages in government organisations, a lack of resources, a desire for independent testing and potential price advantages.
• Supply of AI assurance in D&S is limited by information asymmetries, skills gaps, unclear regulatory guidance and a lack of long-term funding. Demand for AI assurance in D&S is constrained by confusion over assurance offerings, information-sharing barriers, cultural barriers, a lack of funding and slow procurement processes.
• D&S provides a case study with broader lessons for the UK as it works to bolster its AI assurance market. This includes the need to: articulate sector-specific requirements; cultivate a market that caters for different levels of AI assurance maturity; develop initiatives to upskill key stakeholders; create mechanisms to disseminate best practice; and establish certification schemes for AI assurance providers
Automated knowledge discovery for reaction engineering
No full textAccurate, interpretable kinetic models are essential for designing, optimizing, and controlling chemical processes. Traditional mechanistic modeling demands domain expertise, whereas purely data-driven approaches often lack transparency and physical consistency. This thesis presents automated knowledge discovery frameworks that address this by combining symbolic regression, information-theoretic model selection, and physicochemical constraints. Symbolic regression denotes machine learning methods that infer expressions directly from data, rather than assuming a fixed model form, aiming to extract plausible kinetic models with minimal expert intervention.
Chapter 3 examines model selection. A range of information criteria -- including Akaike, Bayesian, generalizable, and Hannan-Quinn -- are assessed to discriminate among kinetic models under noise, limited data, and data richness. A case study illustrates their practical behavior, and model-based design of experiments for model discrimination is introduced.
Chapter 4 introduces two symbolic regression frameworks, ADoK-S and ADoK-W (Automated Discovery of Kinetics, strong and weak formulations), which automate the discovery of rate laws from rate or concentration data. Both rely on genetic programming for expression generation and information-criterion-based refinement, and recover ground-truth models from sparse, noisy datasets.
Chapter 5 extends symbolic regression to mechanism discovery via SiMBA (Simplest Mechanism Builder Algorithm), a four-stage pipeline for mechanism generation, model translation, parameter estimation, and information-criterion-based model comparison. Parallelized backtracking and matrix representations enable efficient exploration of chemically feasible networks, as demonstrated on aldol condensation and fructose dehydration.
Finally, Chapter 6 proposes PI-ADoK (Physics-Informed Automated Discovery of Kinetics), which embeds physicochemical constraints into symbolic regression and incorporates uncertainty quantification via Metropolis-Hastings. Across several catalytic benchmarks, PI-ADoK improves model fidelity, robustness, and data efficiency. Collectively, these contributions advance automated kinetic model discovery for reaction engineering and process systems applications.Open Acces
Complement and neutrophils are actively involved at the cervicovaginal interface in cases of adverse microbial composition, cervical shortening and spontaneous preterm birth
No full textMicrobial-driven spontaneous preterm birth (sPTB) is linked to adverse vaginal microbiome and dysregulated immune responses, yet this knowledge has not translated into predictive tools or therapies. We sampled 186 high-risk pregnant women and assessed 14 complement proteins and the neutrophil immunophenotype at the cervicovaginal interface to expand potential targets. Alterations in classical and alternative complement pathway components were associated with bacterial community composition and preceded cervical shortening and sPTB. At 12⁺⁰–16⁺⁶ weeks, women with Community State Type (CST) IV had significantly higher concentrations of C1q, C4, C4b, Factor B,D, C3b/iC3b, and Factor H,I than those with CST I. Women who later
developed a short cervix and delivered preterm showed lower L. crispatus abundance and elevated complement activation. Neutrophils were the dominant local immune cell and exhibited enhanced activation relative to peripheral neutrophils, with altered expression of CD11b, CD62L, CD63, and CD66b. Cervical neutrophil CD63, CD66b, and CD88 differed between CST IV and CST I, though not by pregnancy outcome. Neutrophil abundance correlated with cytokines, complement proteins, and MMPs, suggesting roles in inflammation and tissue remodelling. These findings highlight a microbiota-driven complement–neutrophil axis present before cervical remodelling and sPTB, identifying potential complement-based predictors and therapeutic targets
Retrotransposable element co-option in the evolution of immune networks
No full textRetrotransposable elements (RTEs) are, or derive from, retrovirus-like genomic parasites which represent almost half of human DNA and can be exonised to provide coding material. In research they have largely been ignored as their repetitive nature creates a significant technical challenge. However, RTEs can provide a powerful source of genetic variation. Indeed, several show evidence for positive selection many millions of years after their integration, suggesting important functions. Here, we have studied the effect of RTE integrations adjacent to the IL13RA1 gene, which encodes a central receptor subunit in type 2 immunity. Two of these integrations (LOR1a and L1MD1) have led to the production of an IL13RA1 transcript with an alternative terminal exon, referred to here as IL13RA1-LOR1a. In the primate genome lineage that contains LOR1a elements, all showed evidence for retention of important motifs necessary for IL13RA1-LOR1a transcript production, and primate lines in culture were positive for expression. As a terminal exon replacement, IL13RA1-LOR1a was subject to a different profile of 3′UTR-mediated regulation, for instance it was resistant to the RNA degrading protein tristetraprolin, which is known to target IL13RA1. In addition, canonical IL13RA1 was found to be sensitive to downregulation by oestrogen, while IL13RA1-LOR1a was not. As a protein, IL-13Rα1-LOR1a’s ligand-binding and transmembrane domains were entirely intact, but its cytoplasmic signalling domain was altered, and experimentally it was signalling defective. When co-expressed, IL-13Rα1-LOR1a was dominant negative over canonical IL-13Rα1, downregulating cellular responses to type 2 immune cytokines. These data therefore suggest balancing the ratio of IL13RA1 isoforms informs overall cellular type 2 immune signalling responses. It is also another example of how RTEs can significantly contribute to evolving immune function.Open Acces