Craig Glastonbury
Research Group Leader
Research Interests
Make your next move.
Prepare your application
Review drafts before you send them.
Explore related searches
Articles (5)
Single-cell transcriptomic and proteomic technologies enable molecular profiling of immune cells at scale but provide limited access to cellular phenotypes shaped by spatial organisation, organelle architecture and cytoskeletal remodelling. Here we present TGlow, a scalable high-content imaging platform optimized for systematic single-cell phenotyping of primary human lymphocytes. TGlow integrates cyclic immunofluorescence, deep z-stack confocal imaging, and open-source data processing pipelines, including both classical and self-supervised vision transformer-based feature extraction, to jointly quantify cellular morphology, organelle organization, and immune activation states. Applied across over 400,000 primary human T cells spanning CD4 + activation time courses, drug perturbations, CRISPR knockouts and CD8 + T-cell exhaustion, TGlow resolves distinct and reproducible phenotypic states. We uncover dose-dependent and mechanism-specific drug phenotypes, such as defective endoplasmic reticulum polarisation under mycophenolic acid and tofacitinib. We show that mitochondrial clustering reveals activation– and cell-cycle–linked remodelling programs, CRISPR perturbations map gene-specific phenotypes that reposition cells along activation trajectories, and we identify a previously unrecognised collapse of cytoskeletal architecture in exhausted CD8 + T cells. TGlow provides a scalable framework for high-dimensional phenotyping of lymphocyte states advancing functional genomics, perturbation screening and population-level immune profiling by resolving the morphological and functional heterogeneity of lymphocytes and enabling systematic linkage of genetic and pharmacological perturbations to cellular function.
Year:
2026
Placenta pathology provides diagnostic insights for understanding pregnancy complications and guides maternal and perinatal care. While placental abnormalities reflect both acute and chronic maternal-fetal health states, the organ’s spatial and temporal heterogeneity poses significant challenges for systematic histological analysis. To date, there has been no analysis of cell populations and their relationship to placental lesions due to the scale of annotations required by expert pathologists. The Histology Analysis Pipeline.PY (HAPPY), previously published by our group, is a hierarchical deep learning approach for quantitative single-cell resolution analysis of H&E slides. Here we apply HAPPY to analyse 130 slides from 62 live full-term pregnancies across healthy controls and four common placental lesion types (infarction, perivillous fibrin, avascular villi, and intervillous thrombosis). We computed cell-type and tissue-structure compositions, established the expected range of healthy variation, and quantified slide-level deviation from this reference using compositional data analysis. Our results reveal significant cellular composition differences between histologically normal and lesioned placentas, including increased extra-villous trophoblasts and leukocytes, coupled with decreases in Hofbauer cells. These changes accompanied distinctive tissue microstructural alterations, particularly increased fibrin deposition and changes to the villous structures. The magnitude of compositional deviation increased with infarction size but not with intervillous thrombosis. Importantly, many differences extend beyond visibly affected areas, indicating organ-wide adaptive responses rather than purely discrete focal pathologies. This quantitative characterisation provides insights into relationships between specific pathologies and placental structure, demonstrating the potential of AI-based methods to enhance conventional histopathological assessment for research and clinical practice.
Year:
2025
Intraplaque haemorrhage (IPH) represents a critical feature of plaque vulnerability as it is robustly associated with adverse cardiovascular events, including stroke and myocardial infarction. How IPH drives plaque instability is unknown. However, its identification and quantification in atherosclerotic plaques is currently performed manually, with high inter-observer variability, limiting its accurate assessment in large cohorts. Leveraging the Athero-Express biobank, an ongoing study comprising a comprehensive dataset of histological, transcriptional, and clinical information from 2,595 carotid endarterectomy patients, we developed an attention-based additive multiple instance learning (MIL) framework to automate the detection and quantification of IPH across whole-slide images of nine distinct histological stains. We demonstrate that routinely available Haematoxylin and Eosin (H&E) staining outperformed all other plaque relevant Immunohistochemistry (IHC) stains tested (AUROC = 0.86), underscoring its utility in quantifying IPH. When combining stains through ensemble models, we see that H&E + CD68 (a macrophage marker) as well as H&E + Verhoeff-Van Gieson elastic fibers staining (EVG) leads to a substantial improvement (AUROC = 0.92). Using our model, we could derive IPH area from the MIL-derived patch-level attention scores, enabling not only classification but precise localisation and quantification of IPH area in each plaque, facilitating downstream analyses of its association and cellular composition with clinical outcomes. By doing so, we demonstrate that IPH presence and area are the most significant predictors of both preoperative symptom presentation and major adverse cardiovascular events (MACE), outperforming manual scoring methods. Automating IPH detection also allowed us to characterise IPH on a molecular level at scale. Pairing IPH measurements with single-cell transcriptomic analyses revealed key molecular pathways involved in IPH, including TNF-α signalling, extracellular matrix remodelling and the presence of foam cells. This study represents the largest effort in the cardiovascular field to integrate digital pathology, machine learning, and molecular data to predict and characterize IPH which leads to better understanding how it drives symptoms and MACE. Our model provides a scalable, interpretable, and reproducible method for plaque phenotyping, enabling the derivation of plaque phenotypes for predictive modelling of MACE outcomes.
Year:
2025
As vast histological archives are digitised, there is a pressing need to be able to associate specific tissue substructures and incident pathology to disease outcomes without arduous annotation. Here, we learn self-supervised representations using a Vision Transformer, trained on 1.7 M histology images across 23 healthy tissues in 838 donors from the Genotype Tissue Expression consortium (GTEx). Using these representations, we can automatically segment tissues into their constituent tissue substructures and pathology proportions across thousands of whole slide images, outperforming other self-supervised methods (43% increase in silhouette score). Additionally, we can detect and quantify histological pathologies present, such as arterial calcification (AUROC = 0.93) and identify missing calcification diagnoses. Finally, to link gene expression to tissue morphology, we introduce RNAPath, a set of models trained on 23 tissue types that can predict and spatially localise individual RNA expression levels directly from H&E histology (mean genes significantly regressed = 5156, FDR 1%). We validate RNAPath spatial predictions with matched ground truth immunohistochemistry for several well characterised control genes, recapitulating their known spatial specificity. Together, these results demonstrate how self-supervised machine learning when applied to vast histological archives allows researchers to answer questions about tissue pathology, its spatial organisation and the interplay between morphological tissue variability and gene expression.
Year:
2024
Biobank-scale imaging provides an unprecedented opportunity to characterise thousands of organ phenotypes, how they vary in populations and how they relate to disease outcomes. However, deriving specific phenotypes from imaging data, such as Magnetic Resonance Imaging (MRI), requires time-consuming expert annotation, limiting scalability, and does not exploit how information-dense such image acquisitions are. In this study, we developed a 3D diffusion autoencoder to derive latent phenotypes from temporally resolved cardiac MRI data of 71,021 UK Biobank participants. These phenotypes were reproducible, heritable ( h 2 = [4 - 18%]), and significantly associated with cardiometabolic traits and outcomes, including atrial fibrillation ( P = 8.5 × 10 −29 ) and myocardial infarction ( P = 3.7 × 10 −12 ). By using latent space manipulation techniques, we were able to learn, directly interpret and visualise what specific latent phenotypes are capturing in a given MRI. To establish the genetic basis of such traits, we performed a genome-wide association study, identifying 89 significant common variants ( P 0.8) linked variants across phenotypic scales, from intermediate cardiac traits to cardiac disease endpoints. For example, rs142556838 that falls in CCDC141 colocalises with a latent imaging phenotype and a diastolic blood pressure locus. Using single-cell RNA-sequencing data we map CCDC141 expression specifically to a population of ventricular cardiomyocytes. Finally, Polygenic Risk Scores (PRS) derived from latent phenotypes demonstrated predictive power for a range of cardiometabolic diseases and enabled us to successfully stratify the individuals into different risk groups. In conclusion, this study showcases the use of diffusion autoencoding methods as powerful tools for unsupervised phenotyping, genetic discovery and disease risk prediction using cardiac MRI data.
Year:
2024
Collaborators (19)
Dominique de Kleijn
Professor
UMC Utrecht
Cecilia Lindgren
Steering Group Member
University of Oxford
Hagit Hochner
Hebrew University of Jerusalem
Michal Mokry
Universitair Medisch Centrum Utrecht
Sander W. van der Laan
Assistant Professor
University Medical Center Groningen
Clint Miller
Associate Professor
Stanford University
Anke Husmann
Assistant Research Professor
University of Cambridge
Andrea Johanna Manrique Rincón
Staff Scientist
Wellcome Sanger Institute
Alessia Visconti
Assistant Professor
University of Turin
Triin Laisk
Associate Professor
University of Tartu
Melina Claussnitzer
Assistant Professor
Harvard University
Adam Levine
NIHR Academic Clinical Lecturer
University College London
Ernest Diez Benavente
Assistant Professor
University Medical Center Groningen
Emanuele Di Angelantonio
University of Cambridge
Francesca Ieva
Associate Professor
Politecnico di Milano
Christoffer Nellåker
Associate Professor
University of Oxford
Gosia Trynka
Science Director and Immune Genomics Group Leader
Wellcome Sanger Institute
Hester den Ruijter
Professor of Cardiovascular Disease in Women
University Medical Center Groningen
Linda M. Ernst
NorthShore University HealthSystem
Make your next move.
Prepare your application
Review drafts before you send them.