Biology subjects

Henriques, R.

Publications and source records attributed to Henriques, R..

At least 19 recordsLinked to original sources

bioRxiv2026

Mapping bioimage-analysis software adoption in microscopy

Most quantitative microscopy measurements pass through image-analysis software, and different programs can return different measurements from the same images. Which programs researchers rely on therefore matters for reproducibility; however, this information lies scattered across the methods sections of the microscopy literature. In this analysis, we show how mentions of 30 software packages changed across 1.7 million full-text microscopy papers published between 2010 and 2025. In 2010, 12% of these papers mentioned at least one of the 22 analysis packages we tracked; by 2025, 30% did. Mentions stayed concentrated on a few general-purpose platforms, led by ImageJ/Fiji. Newer packages are mentioned more often in particular fields, such as QuPath in pathology and napari in electron microscopy. Deep-learning methods are usually mentioned alongside an established platform; papers often run StarDist inside Fiji and Cellpose on its own. Many papers mention a package without citing it, so citation counts understate how often packages appear in published work; authors should make that use visible by citing each package and reporting how they ran it.

bioinformatics↗

bioRxiv2026

AI4Life Open Calls and Public Challenges: why, how, and what we have learned.

Within AI4Life, we ran three Open Calls and three Public Challenges (2023-2025), supporting 22 bioimage analysis projects from 151 applications and engaging 225 challenge participants, with the aim of applying FAIR deep learning in the life sciences. Our experience offers a view of the current state of bioimage analysis, the landscape of available tools, as well as the existing gaps between method developers, tool producers and potential users. It highlights that even after careful selection for AI-ready projects, most still require substantial effort to apply deep learning, and that the field still relies heavily on established, well-rounded methods to solve common problems. We come to the conclusion that for scientific AI in biology, the rate-limiting step is not methods and models but data, annotations, and shared infrastructure underneath them.

bioinformatics↗

bioRxiv2026

NucleiSky enables cross-scale multimodal registration of microscopy data using nuclei constellations

Integrating tissue-level organisation with sub-cellular resolution and molecular information often requires combining multiple microscopy modalities and scales. However, aligning images acquired with different modalities, settings, or instruments remains challenging. Here, we introduce NucleiSky, a microscopy image registration framework that utilises the spatial arrangement of nuclei or other landmarks as an intrinsic biological fingerprint. NucleiSky represents images as constellations of centroids and aligns them using geometric algorithms and spatial consensus scoring. In benchmark datasets, NucleiSky could localise query regions within larger reference images using as few as five nuclei. We show that NucleiSky can locate high-magnification fields of view within low-magnification overview scans, map these alignments to additional channels, support live brightfield-to-fixed registration using synthetic nuclear labels, and guide microscope retargeting. We further show that the same constellation-matching principle can be extended to 3D localisation and to non-nuclear landmarks. These findings establish local landmark geometry as an intrinsic spatial fingerprint that enables localisation and registration across imaging scales, modalities and microscopy platforms. NucleiSky is available as an open-source Python package and as notebook-based applications.

Cell Biology↗

bioRxiv2026

VLab4Mic: prediction of structural resolvability in super-resolution microscopy

Determining whether a microscopy experiment can resolve a specific feature of a protein assembly remains difficult because researchers must balance imaging modality, labelling strategy, and probe choice. We present VLab4Mic, a simulation platform that predicts structural resolvability before experiments. Starting from atomic models from the PDB or AlphaFold predictions, VLab4Mic places antibodies, nanobodies, chemical linkers, or fluorescent proteins on epitopes, applies stochastic labelling and steric constraints, and generates virtual samples for widefield, confocal, AiryScan, Stimulated Emission Depletion (STED), and Single-Molecule Localisation Microscopy (SMLM). Simulations of nuclear pore complexes agree with experimental data across all five modalities. In case studies, HIV capsid appearance depended strongly on orientation, and STED and SMLM separated domed from flat clathrin lattices where confocal and AiryScan did not. VLab4Mic therefore lets researchers judge which biological questions a given imaging configuration can answer before spending time tuning parameters at the microscope.

bioinformatics↗

bioRxiv2026

On why and how to encode probability distributions on graph representations of omics data: enhancing predictive tasks and knowledge discovery

The growing availability and complexity of omics data have driven the development of specialized algorithms for modeling molecular systems. Although graph-based learning methods effectively represent biological interactions, they often neglect the statistical information embedded in node and edge annotations. To address this limitation, we propose a novel graph-based framework that integrates structured statistical distributions into nodes and edges, capturing probabilistic characteristics of molecular relationships. We evaluate the proposed approach on omics datasets from five cancer types across multiple clinical outcomes, including survivability and primary tumor site. Results demonstrate predictive performance comparable to established machine learning baselines. Beyond prediction, the statistically enriched graph representations enable the identification and characterization of regulatory modules associated with clinical outcomes, enhancing biological interpretability. These findings suggest that incorporating structured statistical information into graph representations provides a competitive and interpretable framework for predictive modeling and knowledge discovery in complex diseases.

bioinformatics↗

bioRxiv2025

SMOLM-LFM: ratiometric single molecule orientation without polarizers

Single-molecule orientation and localization microscopy (SMOLM) enables the determination of molecular orientation, wobbling, and position. However, most SMOLM implementations rely on complex point spread function (PSF) fitting, which limits analysis throughput and introduces high computational cost. A way to overcome these limitations is to simplify the analysis using a ratiometric intensity estimation, often relying on polarization projections. While effective in 2D, extending these methods to 3D remains challenging. Here, we introduce a new ratiometric strategy for SMOLM in 3D. Building on the principles of Single Molecule Light Field Microscopy, which captures the 3D position information from a single snapshot by segmenting the back focal plane, we extend this strategy to orientation retrieval. Our approach uses the generalized 3D Stokes formalism to linearly decompose the intensity measurements across the light-field channels, allowing computationally-efficient estimations, while avoiding both complex PSF fitting and polarization projections. This framework, called SMOLM-LFM, enables 6D estimation of single molecules (3D position + 3D orientation) with a simplified optical setup and a large depth-of-field. We present the theoretical foundations, experimental implementation, and validation through measurements on calibration beads, single fluorophores, and cells, thereby demonstrating the methods potential and practical limitations.

bioengineering↗

bioRxiv2025

mAIcrobe: an open-source framework for high-throughput bacterial image analysis

Quantitative analysis in bacterial microscopy is often hindered by diverse cell morphologies, population heterogeneity, and the requirement for specialised computational expertise. To address these challenges, mAIcrobe is introduced as an open-source framework that broadens access to advanced bacterial image analysis by integrating a suite of deep learning models. mAIcrobe incorporates multiple segmentation algorithms, including StarDist, CellPose, and U-Net, alongside comprehensive morphological profiling and an adaptable neural network classifier, all within the napari ecosystem. This unified platform enables the analysis of a wide range of bacterial species, from spherical Staphylococcus aureus to rod-shaped Escherichia coli, across various microscopy modalities within a single environment. The biological utility of mAIcrobe is demonstrated through its application to antibiotic phenotyping in E. coli and the identification of cell cycle defects in S. aureus DnaA mutants. The modular design, supported by Jupyter notebooks, facilitates custom model development and extends AI-driven image analysis capabilities to the broader microbiology community. Building upon the foundation established by eHooke, mAIcrobe represents a substantial advancement in automated and reproducible bacterial microscopy.

microbiology↗

bioRxiv2025

Nanoscale spatial-omics via contrastive embedding of single-molecule localisation data

Omics approaches have revolutionised biology, and cells can now be routinely characterised on the genomic, transcriptomic and proteomic levels. However, there is an additional pillar; the (nanoscale) spatial organisation of molecules in the cell - information now accessible through super-resolution microscopy. We present a contrastive learning framework for nanoscale spatial-omics that embeds single-molecule localisation microscopy data into a latent space representing protein architecture directly to enabling comparative analysis. Using simulated and experimental data, we demonstrate its ability to enable new bioanalysis capabilities including assessing changes to cellular nanoscale architecture arising from pharmacological treatments, cell type, fluorophore selection or data-processing workflows. The approach supports downstream tasks such as clustering proteins by nanoscale organisation, mapping dose-response trajectories and identifying batch effects in replicate datasets, establishing contrastive learning as a scalable foundation for nanoscale spatial-omics and providing a platform for comparative phenotyping, quality control, and hypothesis generation.

bioinformatics↗

bioRxiv2025

Non-synaptic exocytosis along the axon shaft and its regulation by the submembrane periodic skeleton

Neuronal communication relies on signaling molecules transferred via exo- and endocytosis throughout the brain. Historically, studies have focused on vesicle exo- and endocytosis and their release machinery at synapses, and much less is known about non-synaptic exocytosis. If and how vesicles can access the plasma membrane along the axonal shaft, overcoming the insulating layer of the membrane-associated periodic scaffold, remains unclear. Here, we used fast live-cell imaging of mature cultured hippocampal neurons expressing the vamp2-pHluorin reporter to map sponta-neous exocytosis along axons. We detected non-synaptic exocytic events along the axon shaft that concentrated at the axon initial segment. Perturbation of the membrane-associated actin-spectrin skeleton revealed its role in regulating shaft exocytosis, similarly to its recently demonstrated role in shaping axonal endocytosis. To visualize the nanoscale arrangement of exocytic locations, we developed a novel correlative live-cell/two - color, single-molecule localization microscopy (SMLM) approach. We observed that regions of exocytosis are devoid of the submembrane spectrin mesh, with these spectrin-free areas being spatially separated from the spectrin clearings that contain clathrin-coated pits. Overall, our work reveals a new process of spontaneous exocytosis along the axon shaft, and how the axonal submem-brane skeleton shapes a heterogeneous landscape that uniquely segregates vesicular trafficking events.

cell biology↗

bioRxiv2025

Spatiotemporal Coordination of Guidance Cues Directs Multipolar Migration During Retinal Lamination

Multipolar migration is a conserved neuronal migration mode in the developing brain, enabling emerging neurons to navigate crowded environments and reach precise laminar positions. Yet how these cells interpret external cues to guide their migration remains unclear. We investigate this question in the developing vertebrate retina using horizontal cells as a model. Combining transcriptomics, targeted CRISPR screening, and live imaging, we reveal the spatiotemporal guidance system underlying horizontal cell lamination: repulsive Slit1b/2-Robo2 signaling in the amacrine cell layer is essential to initiate apical horizontal cell migration, while attractive Neurturin-Gfr1 signaling from photoreceptors fine-tunes final positioning beneath the photoreceptor layer. Disruption of these pathways causes basal retention of horizontal cells, highlighting the importance of spatially coordinated signaling for proper lamination and functional retinal circuitry. Our results uncover how positional signals and tissue architecture cooperate to achieve neuronal precision, an organizing principle likely relevant across the developing central nervous system.

developmental biology↗

bioRxiv2025

VP-CLEM-Kit: An accessible pipeline for visual proteomics using super resolution volume correlative light and electron microscopy (SR-vCLEM)

Visual proteomics (VP) aims to allow researchers to visualise, measure and analyse proteins in the context of cell and tissue structure in health and disease. VP is becoming a reality through technological advances across several domains, including in situ structural biology and correlative light and electron microscopy (CLEM). However, widespread adoption remains limited due to the complexity and cost of the various VP approaches reported to date. Here we present the VP-CLEM-Kit, a disruptive cost-effective pipeline for super resolution volume CLEM (SR-vCLEM) that can be implemented with minimal advanced electron microscopy expertise and equipment, making it accessible to light microscopy facilities and research labs. SR-vCLEM is based on in-resin fluorescence (IRF), where fluorophores are preserved through processing into resin. The easyIRF protocol reported here reduces the requirement for complex costly sample preparation equipment and toxic chemicals compared to standard IRF protocols. easyIRF blocks are cut into ultrathin sections that are imaged using tomoSTORM, a new modular and cost-effective openFrame-based light microscope controlled by the open-source software package Micro-Manager that provides serial single molecule localisation microscopy in array tomography format. Sections are then post-stained and imaged using a tabletop scanning electron microscope controlled by open-source SBEMimage software to run in array tomography format. We demonstrate the potential of the VP-CLEM-Kit by imaging organelle reporters in human cell lines and stem cell derived neurons, fluorescently labelled protein in neurons, and immunolabelled cells in human kidney biopsy tissue from transplant patients. The VP-CLEM-Kit delivers a [~]5-fold improvement in resolution at a [~]7-fold lower cost, and thus provides new technical capability as well as a blueprint for more equitable access to advanced imaging workflows.

cell biology↗

bioRxiv2025

ReScale4DL: Balancing Pixel and Contextual Information for Enhanced Bioimage Segmentation

Deep learning has established itself as the state-of-the-art approach for segmentation in bioimage analysis. However, these powerful algorithms present an intriguing paradox regarding image resolution: contrary to intuition, lower-resolution images can yield superior performance for specific image analysis carried out by deep learning. This phenomenon is particularly significant in microscopy imaging, where high-resolution acquisitions come with substantial costs in throughput, data storage requirements, and potential photodamage to specimens. Through systematic experimentation, we evaluate how varying image resolution impacts deep learning performance in cellular image segmentation tasks. We trained popular architectures on datasets downsampled to 6-50% of their original resolution, mimicking acquisitions at lower image magnification, and compared their performance against models trained on native-resolution images. Our results show that segmentation accuracy either improves (by up to 25% of mean Intersection over Union (IoU)) or experiences only minimal degradation (< 5% of mean IoU) when using images downsampled by up to a factor of 4 (25% of the original resolution). This downsampling proportionally increases information throughput while reducing data storage requirements and inference time. With these findings, we contribute systematic guidelines to deep learning practitioners in creating efficient experimental pipelines for image-driven discoveries. This approach improves the sustainability and cost-effectiveness of bioimaging studies by reducing data and computing needs while optimising microscopy techniques.

biophysics↗

bioRxiv2024

Structural Repetition Detector: multi-scale quantitative mapping of molecular complexes through microscopy

From molecules to organelles, cells exhibit recurring structural motifs across multiple scales. Understanding these structures provides insights into their functional roles. While super-resolution microscopy can visualise such patterns, manual detection in large datasets is challenging and biased. We present the Structural Repetition Detector (SReD), an unsupervised computational framework that identifies repetitive biological structures by exploiting local texture repetition. SReD formulates structure detection as a similarity-matching problem between local image regions. It detects recurring patterns without prior knowledge or constraints on the imaging modality. We demonstrate SReDs capabilities on various fluorescence microscopy images. Quantitative analyses of three datasets highlight SReDs utility: estimating the periodicity of spectrin rings in neurons, detecting HIV-1 viral assembly, and evaluating microtubule dynamics modulated by EB3. Our open-source ImageJ and Fiji plugin enables unbiased analysis of repetitive structures across imaging modalities in diverse biological contexts.

cell biology↗

bioRxiv2024

Nano-org, a functional resource for single-molecule localisation microscopy data

We present a publicly accessible, curated, and functional resource, termed "nano-org", containing single-molecule localisation microscopy (SMLM) data representing the nanoscale distributions of proteins in cells. Nano-org is searchable by comparing the statistical similarity of the datasets it contains. This unique functionality allows the resource to be used to understand the relationships of nanoscale architectures between proteins, cell types or conditions, enabling a new field of spatial nano-omics.

biophysics↗

bioRxiv2024

PhotoFiTT: A Quantitative Framework for Assessing Phototoxicity in Live-Cell Microscopy Experiments

Phototoxicity in live-cell fluorescence microscopy can compromise experimental outcomes, yet quantitative methods to assess its impact remain limited. Here we present PhotoFiTT (Phototoxicity Fitness Time Trial), an integrated framework combining a standardised experimental protocol with advanced image analysis to quantify light-induced cellular stress in label-free settings. PhotoFiTT leverages machine learning and cell cycle dynamics to analyse mitotic timing, cell size changes, and overall cellular activity in response to controlled light exposure. Using adherent mammalian cells, we demonstrate PhotoFiTTs ability to detect wavelength- and dose-dependent effects, showcasing that near-UV light induces significant mitotic delays at doses as low as 0.6 J/cm2, while longer wavelengths require higher doses for comparable effects. PhotoFiTT enables researchers to establish quantitative benchmarks for acceptable levels of photodamage, facilitating the optimisation of imaging protocols that balance image quality with sample health.

cell biology↗

bioRxiv2023

DL4MicEverywhere: Deep learning for microscopy made flexible, shareable, and reproducible

Deep learning has revolutionised the analysis of extensive microscopy datasets, yet challenges persist in the widespread adoption of these techniques. Many lack access to training data, computing resources, and expertise to develop complex models. We introduce DL4MicEverywhere, advancing our previous ZeroCostDL4Mic platform, to make deep learning more accessible. DL4MicEverywhere uniquely allows flexible training and deployment across diverse computational environments by encapsulating methods in interactive Jupyter notebooks within Docker containers -a standalone virtualisation of required packages and code to reproduce a computational environment-. This enhances reproducibility and convenience. The platform includes twice as many techniques as originally provided by ZeroCostDL4Mic and enables community contributions via automated build pipelines. DL4MicEverywhere empowers participatory innovation and aims to democratise deep learning for bioimage analysis.

bioinformatics↗

bioRxiv2023

Transertion and cell geometry organize the Escherichia coli nucleoid during rapid growth

AO_SCPLOWBSTRACTC_SCPLOWBacterial chromosomes are spatiotemporally organized and sensitive to environmental changes. However, the mechanisms underlying chromosome configuration and reorganization are currently not fully understood. Using 3D single-molecule localization microscopy and live-cell imaging, we show that the Escherichia coli nucleoid adopts a condensed, membrane-associated configuration during rapid growth. To study the influence of different biosynthetic processes on nucleoid morphology and positioning, we recorded multi-colour super-resolution images during drug treatment. After developing analysis routines for confocal and super-resolution images, we captured highly resolved snapshots which revealed the complete loss of the membrane-bound state of the nucleoid within 10 minutes of halting transcription and translation. This indicates an active role of transertion (coupled transcription, translation and membrane insertion) in nucleoid organization. In contrast, cell wall synthesis inhibition only affects nucleoid organization during morphological changes. Further, we provide evidence that the E. coli nucleoid spatially correlates with MreB in unperturbed E. coli cells, while this correlation diminishes in cells with changed cell geometry or upon inhibition of protein biosynthesis. Replication inhibition experiments, as well as multi-drug treatments highlight the role of entropic effects and transcription in nucleoid condensation and positioning. In summary, we provide experimental evidence for transertion as a principal organiser of the bacterial nucleoid, and show that an altered metabolic state and antibiotic treatment both lead to major changes in the degree of transertion and overall spatial organization of the nucleoid. Our high-resolution characterization reveals dynamics of antibiotic action and provides tools to quantify bacterial chromosome organization. This does not only provide valuable insights into the role of transertion, but can also be applied to study other cell-biological processes.

microbiology↗

bioRxiv2023

NanoPyx: super-fast bioimage analysis powered by adaptive machine learning

To overcome the challenges posed by large and complex microscopy datasets, we have developed NanoPyx, an adaptive bioimage analysis framework designed for high-speed processing. At the core of NanoPyx is the Liquid Engine, an agent-based machine-learning system that predicts acceleration strategies for image analysis tasks. Unlike traditional single-algorithm methods, the Liquid Engine generates multiple CPU and GPU code variations using a meta-programming system, creating a competitive environment where different algorithms are benchmarked against each other to achieve optimal performance under the user"s computational environment. In initial experiments focusing on super-resolution analysis methods, the Liquid Engine demonstrated an over 10-fold computational speed improvement by accurately predicting the ideal scenarios to switch between algorithmic implementations. NanoPyx is accessible to users through a Python library, code-free Jupyter notebooks, and a napari plugin, making it suitable for individuals regardless of their coding proficiency. Furthermore, the optimisation principles embodied by the Liquid Engine have broader implications, extending their applicability to various high-performance computing fields.

bioinformatics↗