Most of the members of this group are from the Statistics Section and Biomaths research group of the Department of Mathematics. Below you can find a list of research areas that members of this group are currently working on and/or would like to work on by applying their developed mathematical and statistical methods.
Research areas
Research areas
- Systems Biology
- Statistical genomics and Epidemiology
- Medical Imaging
- Precision and Stratified Medicine
- Analysis of clinical trials, observational and longitudinal studies
- Infectious Disease Epidemiology
Publications
Results
- Showing results for:
- Reset all filters
Search results
-
Journal articleLhoste VPF, Fan Y, Bennett JE, et al., 2026,
Hierarchical Bayesian mixture model for subgroup-specific cardiometabolic phenotypes
, Journal of the Royal Statistical Society Series A: Statistics in Society, ISSN: 0964-1998<jats:title>Abstract</jats:title> <jats:p>Anthropometric, cardiometabolic, and renal traits—including height, adiposity, blood pressure, blood glucose, cholesterol, and kidney function—affect numerous noncommunicable and infectious diseases. These traits share some genetic, environmental and nutritional determinants while also having distinct causes. Their complex distributions and correlations further vary among population subgroups. Traditional clustering methods for multiple continuous variables often overlook subgroup-specific distributions or hinder comparisons across separately clustered subgroups. We developed a Bayesian hierarchical mixture model to coherently identify anthropometric, cardiometabolic and renal phenotypes across population subgroups. Our model balances flexibility and comparability by adapting a Bayesian hierarchical mixture model to include subgroup-specific cluster weights and empirically driven shifts in cluster means. This structure allows subgroup-specific variations on phenotype characteristics while maintaining a shared structure and interpretation. We applied this model to anthropometric, cardiometabolic and renal data of men from three different age subgroups in the US National Health and Nutrition Examination Surveys. The results demonstrate the model’s ability to identify relevant phenotypes within each age subgroup and maintain comparability across them. The model can be applied to a range of other unsupervised clustering tasks related to health, sociodemographic or environmental data that have complex patterns with similarities and differences across subgroups.</jats:p>
-
Journal articleHalder A, Peiris HV, Thorp S, et al., 2026,
pop-cosmos: redshifts and physical properties of KiDS-1000 galaxies
, Monthly Notices of the Royal Astronomical Society, Vol: 551, ISSN: 0035-8711Principled Bayesian inference of galaxy properties has not previously been performed for wide-area weak-lensing surveys with millions of sources. We address this gap by applying the pop-cosmos generative model to perform spectral energy distribution (SED) fitting for 4 million KiDS (Kilo-Degree Survey)-1000 galaxies. Calibrated on deep COSMOS2020 photometric data, pop-cosmos specifies a physically motivated prior over the galaxy population up to (Formula presented) in stellar population synthesis (SPS) parameter space. Using the Speculator SPS emulator with GPU (graphics processing unit)-accelerated Markov Chain Monte Carlo sampling, we perform full posterior inference at 8.2 GPU seconds per galaxy, obtaining joint constraints on galaxy redshifts and physical properties. We validate photometric redshifts against (Formula presented) KiDS galaxies cross-matched to Dark Energy Spectroscopic Instrument Data Release 1 spectroscopic samples, achieving low bias ((Formula presented) ), scatter ((Formula presented) ), and outlier fraction (3.2 per cent) for the Bright Galaxy Survey, with comparable performance (bias (Formula presented), (Formula presented), 1.0 per cent outliers) for luminous red galaxies (LRGs). Within the LRG sample, we identify massive, dusty, star-forming contaminants at (Formula presented) satisfying standard colour selections for quenched populations. We infer trends in stellar mass, star formation, metallicity, and dust across five tomographic redshift bins consistent with established scaling relations. Using specific star formation rate constraints, we identify (Formula presented) 7 per cent of KiDS-1000 galaxies as quenched, versus 37 per cent implied by conservative colour cuts. This enables the construction of weak-lensing samples defined by physical properties while mitigating intrinsic alignment systematics and preserving statistical power. Our analysis validates pop-cosmos out of sample, esta
-
Journal articleKaveh F, Green A, Jones N, 2026,
The Joint Spectrum over Trees under the Kingman coalescent with varying population.
, Theor Popul Biol, Vol: 171, Pages: 59-78The Kingman coalescent process models the genealogy of a sample taken from a large population of individuals who reproduce and die according to models such as the Moran or Wright-Fisher processes. The occurrence and spread of neutral mutations in the sample can be modelled by a Poisson process over sample phylogenies from the Kingman coalescent. We study the joint probability distribution of frequencies for multiple mutations occurring on the same tree. We call this the Joint Spectrum over Trees (JST). We derive a closed-form solution for this joint distribution in the case of two mutations with varying population size. We specialise the result for specific population histories, including constant population size. In the process, we highlight how different averaging procedures can lead to different distributions for the frequency of mutations, even when considering only the frequency of a single mutation. We provide a systematic approximation scheme for the Joint Spectrum over Trees under constant population when the number of samples is large. The exact form of the Joint Spectrum over Trees has implications for genealogical inference with the Kingman coalescent, specifically for the characterisation of tree structure near the root and in parameter inference when the underlying tree structures are unknown. To this end, we also comment on the validity of the independence approximation to the true joint distribution under different population histories.
-
Journal articleChertock A, Degond P, Sagiv A, et al., 2026,
The Evolution of Pointwise Statistics in Hyperbolic Equations with Random Data
, SIAM/ASA Journal on Uncertainty Quantification, Vol: 14, Pages: 1045-1079<jats:p>Abstract.</jats:p> <jats:p>We consider one-dimensional hyperbolic PDEs, linear and nonlinear, with random initial data. Our focus is the pointwise statistics, i.e., the probability measure of the solution at any fixed point in space and time. For linear hyperbolic equations, the probability density function (PDF) of these statistics satisfies the same linear PDE. For nonlinear hyperbolic PDEs, we derive a linear transport equation for the cumulative distribution function (CDF) and a nonlocal linear PDE for the PDF. Both results are valid only as long as no shocks have formed, a limitation which is inherent to the problem, as demonstrated by a counterexample. For systems of linear hyperbolic equations, we introduce the multi-point statistics and derive their evolution equations. In all of the settings we consider, the resulting PDEs for the statistics are of practical significance: they enable efficient evaluation of the random dynamics, without requiring an ensemble of solutions of the underlying PDE, and their cost is not affected by the dimension of the random parameter space. Additionally, the evolution equations for the statistics lead to a priori statistical error bounds for Monte Carlo methods (in particular, kernel density estimators) when applied to hyperbolic PDEs with random data.</jats:p>
-
Journal articleBeaney T, Clarke J, Woodcock T, et al., 2026,
Challenges of clustering disease trajectories in people with Multiple Long-Term Conditions using data from 7.2 million patients
, Communications Medicine, ISSN: 2730-664XBackgroundGenerating clusters of people with similar patterns of Multiple Long-Term Conditions (MLTC) could help target healthcare services for specific patient groups. We aimed to generate data-driven clusters of patients based on their disease trajectories and assess clinical interpretability and associations with future health outcomes.MethodsWe used structured general practice data from the Clinical Practice Research Datalink Aurum, linked to Hospital Episode Statistics data on emergency department (ED) attendances and hospital admissions, including all adults registered on 1st January 2015. For each patient, we generated a vector embedding representing their trajectory of diseases developed over time, using a transformer model, followed by clustering using k-means. Patients were subsequently followed up for 1 year to estimate associations of cluster membership with ED attendance, hospitalisation and mortality, compared to patients with no long-term conditions. We also evaluated the use of the clusters for predicting these outcomes compared with using number of long-term conditions, individual diseases or embeddings alone.Results Analysis included 5,981,091 (82.1%) patients with MLTC and 1,304,119 (17.9%) with no chronic conditions. We identified eight clusters of patients with MLTC as optimal. The largest, representing 21.3% of the population included strong contributions from cardio-kidney-metabolic conditions, but there was substantial overlap in prevalent conditions across clusters. Although large differences were found between clusters in 1-year odds of ED attendance, hospitalisation and mortality, with the highest odds in cardio-kidney-metabolic clusters, clusters explained only between 1.2-3.1% of total variance, and associations were substantially attenuated after adjustment for age, sex, ethnicity and deprivation. Clusters performed substantially worse at predicting outcomes than using the individual diseases or embeddings, and performed worse than a pe
-
Journal articleBaum M, Sanna Passino F, Gandy A, 2026,
Statistical hypothesis testing for differences between layers in dynamic multiplex graphs
, IEEE Transactions on Signal and Information Processing over Networks, ISSN: 2373-7778 -
Journal articleNg WK, Dranczewski J, Fischer A, et al., 2026,
Few-shot neuromorphic vision in a nonlinear photonic network laser
, Science Advances, Vol: 12, ISSN: 2375-2548With the growing prevalence of AI, demand increases for hardware that mimics the brain’s ability to extract structure from limited data. In the retina, ganglion cells detect features from sparse inputs via lateral inhibition, where neurons antagonistically suppress activity of neighbouring cells. Biological neurons exhibit diverse heterogeneous nonlinear responses, linked to robustlearning and strong performance in low-data regimes.Here, we introduce a retinally-inspired photonic computing system where spatially-competing lasing modes in a random network laser act as heterogeneous, inhibitively-coupled neurons- enabling feature detection, few-shot classification, and segmentation. This silicon-compatible scheme harnesses heterogeneous excitatory and inhibitory nonlinear physical dynamics which give rise to emergent photonic computing behaviour, including parallel feature detection and strong performance when training data is scarce. We report 98.05% and 87.85% accuracy on MNIST and Fashion-MNIST, and 90.12% on BreaKHis cancer diagnosis- outperforming software CNNs including EfficientNetV2 and the vision transformer ViT in few-shot and class-imbalanced regimes with training sets of up to several hundred images. We demonstrate combined segmentation and classification on the HAM10k skin lesion dataset, achieving DICE and Jaccard scores of 84.49% and 74.80%. These results demonstratethe potential of random lasing networks as nonlinear photonic learning systems, and highlight the ability of heterogeneous nonlinear dynamics to support strong learning in challenging low-data scenarios.
-
Book chapterMichalewicz K, Barahona M, Bravi B, 2026,
Machine Learning Approaches for Interpretable Antibody Property Prediction Using Structural Data
, Artificial Neural Networks, Editors: Cartwright, Publisher: Humana, Pages: 171-196, ISBN: 9781071654071 -
Journal articleTucker S, Baldonado N, Ruina OH, et al., 2026,
Effectiveness of the hope groups intervention on mental health, parenting practices, and violence against children among Ukrainian parents and caregivers affected by war and displacement: a cluster randomised controlled trial
, The Lancet Regional Health - Europe, ISSN: 2666-7762BackgroundAlmost two-thirds of the world’s children live in conflict-affected countries, including Ukraine, where two-thirds of children were displaced after Russia’s 2022 invasion. To address the urgent need for effective approaches to support parents and protect children in war, we designed Hope Groups—a group-based mental health, parenting, and violence prevention intervention—and used a cluster randomised controlled trial (cRCT) to evaluate its effectiveness.MethodsWe conducted a two-arm, parallel cRCT comprising K = 90 clusters and N = 510 parents/caregivers, who were internally displaced, externally displaced, or at home in war-affected Ukraine. Primary outcomes, assessed at baseline and after completing the 12-session intervention (endline), included caregiver mental health, parenting practices, and violence against children. Clusters were randomly assigned to immediate Hope Groups intervention (K = 45; N = 255) or waitlist control (K = 45; N = 255). Interim analyses were conducted after the first 20 clusters reached trial endline. The trial was prospectively registered with Open Science Framework (https://osf.io/uvj67) and subsequently registered on ClinicalTrials.gov to align with ICMJE guidance (NCT07470333).FindingsAfter completion of the 12-session intervention, the Hope Groups intervention demonstrated medium-to-large effects across all measured primary outcomes compared with the control arm. Caregivers in the intervention arm reported 57.4% lower levels of parental depression and anxiety (β = −2.85; 95% CI = −3.49, −2.22; d = −0.86); 52.7% lower levels of violence against children (β = −0.5; 95% CI = −0.64, −0.36; d = −0.43); 33% higher parenting practices (β = 1.32, 95% CI = 0.92, 1.71, d = 0.57), compared to the control arm. Response rate was high: 98.6%. There were no adverse events or harms.InterpretationThe results demonstrate that the Hope Groups intervention i
-
Journal articleGeorgiev D, Xie R, Reumann D, et al., 2026,
Label-free biochemical imaging and timepoint analysis of neural organoids via deep learning-enhanced Raman microspectroscopy
, Science Advances, Vol: 12, ISSN: 2375-2548Three-dimensional (3D) organoids have emerged as powerful models for studying human development, disease, and drug response in vitro. Yet, their analysis remains constrained by standard imaging and characterization techniques, which are invasive, require exogenous labeling, and offer limited multiplexing. Here, we present a noninvasive, label-free imaging platform that integrates Raman microspectroscopy with deep learning–based hyperspectral unmixing for unsupervised, spatially resolved biochemical analysis of neural organoids. Our approach enables 2D and 3D mapping of cellular and subcellular structures in both cryosectioned and intact organoids, achieving improved imaging accuracy and robustness compared to conventional methods for hyperspectral analysis. Using our platform, we demonstrate volumetric imaging of a neural rosette within a neural organoid and interrogate changes in biochemical composition during early developmental stages in intact neural organoids, revealing spatiotemporal variations in lipids, proteins, and nucleic acids. This work establishes a versatile framework for high-content, label-free (bio)chemical phenotyping with broad applications in organoid research and beyond.
This data is extracted from the Web of Science and reproduced under a licence from Thomson Reuters. You may not copy or re-distribute this data in whole or in part without the written consent of the Science business of Thomson Reuters.
Contact us
If you are interested in meeting with members of the group please contact Marina Evangelou