eLife: latest articles by subject

Role of desolvation on biomolecular liquid–liquid phase separation

wfli@nju.edu.cn (Kai Zhang) — Fri, 09 Oct 2026 00:00:00 +0000

Biomolecular condensates play essential roles in cellular organization and are implicated in diverse pathological processes. Their formation is driven by liquid–liquid phase separation (LLPS), a process that requires coordinated multistep desolvation of biomolecular chains and multivalent inter-chain interactions. Although coarse-grained (CG) models with implicit solvent are widely used to probe LLPS thermodynamics and kinetics, they typically neglect water-mediated desolvation effects, limiting their accuracy and mechanistic interpretability. Here, guided by all-atom simulations and experimental measurements, we develop a desolvation-aware implicit-solvent CG model by incorporating residue-level desolvation terms directly into the pairwise energy function, and apply it to investigate LLPS of intrinsically disordered proteins. Incorporating these desolvation interactions reshapes the phase diagram, alleviating dense-phase overcompaction. Notably, we observe an approximately linear correlation between the temperature gap (simulation temperature relative to the critical point) and the extent of conformational expansion accompanying the dilute-to-dense phase transition, a result further supported by theoretical analysis. We also find that desolvation barriers slow early density-fluctuation growth and shorten transient kinetic arrest, whereas solvent-separated contact interactions exert the opposite effects. Both terms further modulate chain mobility within mature condensates through competing packing and energy-landscape effects. Together, this framework enables an efficient representation of desolvation in CG simulations and reveals how desolvation energetics shape both the thermodynamic landscape and kinetic properties of biomolecular LLPS.

A real-time, multi-animal model for automatic face detection and identification of freely moving common marmosets based on YOLOv8 algorithms

jiayue.yang@mail.mcgill.ca (James Wang) — Tue, 29 Sep 2026 00:00:00 +0000

Precise and up-to-date information about animal location and identity allows us to better quantify individual behaviors in studies of neural activity, cognition, and animal health. In socially housed laboratory animals, identification is usually defined by observation or invasive markers, making the data collection time-consuming, variable across experimenters, and disruptive to animals. We established an automatic pipeline for real-time identification of common marmosets in captivity using a close-view camera. It uses the supervised deep-learning YOLOv8 model to localize individuals, detect faces, and classify identities. Moreover, we use recognition of uniquely color-coded collar beads to improve detection accuracy among visually similar individuals. Across adult and juvenile marmosets, our system automatically identifies marmosets with >82.9% precision and >91.5% recall, achieving human-level performance. This pipeline is designed to be easy to use and generalizable across non-human primate species, ages, and recording hardware, providing rapid and automatic identity recognition from real-time video.

Single-cell spatial mapping reveals reproducible cell type organization and spatially dependent gene expression in gastruloids

arjunrajlab@gmail.com (Arjun Raj) — Tue, 29 Sep 2026 00:00:00 +0000

Gastruloids are stem-cell-based models that recapitulate key aspects of mammalian gastrulation, including the formation of an anterior-posterior axis. However, we do not have detailed spatial information about gene expression and cell type organization, particularly at the level of individual gastruloids. Here, we report a spatially resolved, single-cell molecular catalog of the transcriptomes of 26 individual gastruloids. We found that cell type composition and tissue-scale spatial organization were largely consistent across gastruloids, but meso-scale patterning of specific cell types varied between samples. Posterior cell types formed distinct, organized clusters, while anterior cell types were more disorganized. To distinguish progressive differentiation from cell type differences, we developed the L-score, a parameter-free quantification of mutually exclusive gene expression. This analysis revealed spatial organization without explicit encoding, recapitulated known cell type relationships, and identified novel gene expression states and spatial subclusters within cell types. We confirmed that in gastruloids, neuromesodermal precursor differentiation occurred through a continuous, spatially coordinated process. We also showed that endothelial precursors exhibited unique spatial organization and had distinct gene expression profiles dependent on their association with anterior somitic or posterior endodermal tissues. This work enables the rigorous use of gastruloids as models for studying the molecular mechanisms underlying mammalian development and tissue organization and introduces new computational tools for analyzing spatially resolved single-cell datasets.

Opening the black box toward a modular approach to spike sorting

samuel.garcia@cnrs.fr (Alessio Paolo Buccino) — Fri, 25 Sep 2026 00:00:00 +0000

Spike sorting is an algorithmic process that extracts the activity of individual neurons from extracellular electrophysiology recordings. With the ballooning use of high-density probes, such as Neuropixels, this essential processing step is increasingly becoming time-consuming and computationally expensive. Although many software tools have been proposed to address spike sorting, they are usually constructed and benchmarked as monolithic ‘black boxes’, making it difficult to factor out the effects of individual algorithmic steps on the final outcome, especially when varying datasets and parameters. To address this issue, we developed a modular and common framework to develop, benchmark, and assemble the key computational steps that are used in state-of-the-art spike sorting algorithms. Relying on fast and efficient ground truth generation of biophysically plausible recordings, we show that we are able to individually benchmark and precisely quantify the performance of different steps in a spike sorting pipeline (i.e. peak detection, feature extraction, clustering, and template matching). We then leverage these results to create a modular, component-based spike sorter that can outperform Kilosort4 on dense and large simulated recordings, and produce similar quantitative results on real data. In addition, we find that the major bottleneck of all modern spike sorting pipelines is in the physical motion of probes, regardless of the drift-correction strategy. The component-based spike sorting framework presented here has the potential to foster community engagement in the field by lowering the barrier to contributions and providing a flexible yet powerful framework to construct end-to-end spike sorting solutions.

Real-time closed-loop feedback system for mouse mesoscale cortical signal and movement control

thmurphy@mail.ubc.ca (Pankaj Kumar Gupta) — Thu, 24 Sep 2026 00:00:00 +0000

Increasingly, experiments designed to provide practical perturbations to circuits or behavior are required for hypothesis testing in various disciplines ranging from motor learning to recovery after injury. We present the implementation and efficacy of an open-source closed-loop neurofeedback (CLNF) and closed-loop movement feedback (CLMF) system. In CLNF, we measure mm-scale cortical mesoscale activity with GCaMP6s and provide graded auditory feedback (within ~63 ms) based on changes in dorsal-cortical activation within regions of interest (ROIs) and with a specified rule. Single or dual ROIs (ROI1, ROI2) on the dorsal cortical map were selected as targets. Both motor and sensory regions supported closed-loop training in male and female mice. Mice modulated activity in rule-specific target cortical ROIs to get increasing rewards over days (repeated-measures ANOVA [RM-ANOVA], p=2.83e-5) and adapted to changes in ROI rules (RM-ANOVA, p=8.3e-10, Table 4 for different rule changes). In CLMF, feedback (within ~67 ms) was based on tracking a specified body movement, and rewards were generated when the behavior reached a threshold. For movement training, the group that received graded auditory feedback performed significantly better (RM-ANOVA, p=9.6e-7) than a control group (RM-ANOVA, p=0.49) within 4 training days. Additionally, mice can learn a change in task rule from left forelimb to right forelimb within a day, after a brief performance drop on day 5. Offline analysis of neural data and behavioral tracking revealed changes in the overall distribution of Ca²⁺ fluorescence values in CLNF and body-part speed values in CLMF experiments. Increased CLMF performance was accompanied by a decrease in task latency and cortical ΔF/F₀ amplitude during the task, indicating lower cortical activation as the task gets more familiar.

Linear antibody epitope prediction using AlphaFold2

christopher.snow@colostate.edu (Brian J Geiss) — Fri, 18 Sep 2026 00:00:00 +0000

Defining the binding epitopes of antibodies is essential for understanding how they bind to their antigens and perform their molecular functions. However, while determining linear epitopes of monoclonal antibodies can be accomplished utilizing well-established empirical procedures, these approaches are generally labor- and time-intensive, and costly. To take advantage of the recent advances in protein structure prediction algorithms available to the scientific community, we developed a calculation pipeline based on the localColabFold implementation of AlphaFold2 that can predict linear antibody epitopes by predicting the structure of the complex between antibody heavy and light chains and target peptide sequences derived from antigens. We found that this AlphaFold2 pipeline, which we call PAbFold, was able to accurately flag known epitope sequences for several well-known antibody targets (HA/Myc) when the target sequence was broken into small overlapping linear peptides and antibody complementarity determining regions were grafted onto several different antibody framework regions in the single-chain antibody fragment format. To determine if this pipeline was able to identify the epitope of a novel antibody with no structural information publicly available, we determined the epitope of a novel anti-SARS-CoV-2 nucleocapsid-targeted antibody using our method and then experimentally validated our computational results using peptide competition ELISA assays. These results indicate that the AlphaFold2-based PAbFold pipeline we developed is capable of accurately identifying linear antibody epitopes in a short time using just antibody and target protein sequences. This emergent capability of the method is sensitive to methodological details such as peptide length, AlphaFold2 neural network versions, and multiple-sequence alignment databases. PAbFold is available at https://github.com/jbderoo/PAbFold.

The power of theory in the life sciences

jacob.fine@mail.utoronto.ca (Adam MR Groh) — Fri, 18 Sep 2026 00:00:00 +0000

The rapid growth of high-throughput biology and genomics over the past two decades has helped catalogue many different aspects of gene function in diverse cell types and conditions. More recently, advances in artificial intelligence and deep learning have shown tremendous promise in making accurate predictions of functional genomics measurements. These advances make it tempting to equate experimental cataloguing and accurate prediction with the growth of our theoretical understanding of biological processes – an equivalence we believe is ultimately misleading.

Comprehensive RNA velocity by modeling the cascade of gene regulation, transcription, and splicing from single-cell RNA sequencing data with TSvelo

yyuan@ipe.ac.cn (Hong-Bin Shen) — Tue, 15 Sep 2026 00:00:00 +0000

RNA velocity approaches fit gene dynamics and infer cell fate by modeling the splicing process using single-cell RNA sequencing (scRNA-seq) data. However, due to the short time scale of splicing, high noise, and large complexity of data, existing RNA velocity methods often fail to precisely capture the complex velocity dynamics for individual genes and single cells, which makes their downstream analysis less reliable and less robust. We propose TSvelo, a comprehensive RNA velocity mathematics framework that can model the cascade of gene regulation, Transcription and Splicing using highly interpretable neural ordinary differential equations. TSvelo can precisely capture the transcription–unspliced–spliced 3D dynamics of all genes simultaneously, infer unified latent time shared by genes within a single cell, and be applied to multi-lineage datasets. Experiments on six scRNA-seq datasets, including two multi-lineage datasets, demonstrate TSvelo’s superiority.

+Clonal stochasticity in early NK cell response to mouse cytomegalovirus is generated by mature subsets of varying proliferative ability

darren.wethington@nationwidechildrens.org (Darren Wethington) — Mon, 14 Sep 2026 00:00:00 +0000

Natural killer (NK) cells are classically defined as innate immune cells, but experiments show that mouse cytomegalovirus (MCMV) infection in C57BL/6 mice can cause NK cells to undergo antigen-specific proliferation and memory formation, similar to adaptive CD8+ T cells. One shared behavior between CD8+ T cells and NK cells is clonal expansion, where a single stimulated cell proliferates rapidly to form a diverse population of cells. For example, clones derived from single cells are most abundant during expansion when they are primarily CD27- for NK cells and CD62L- for T cells, phenotypes derived from precursor CD27+ and CD62L + cells, respectively. Here we determined the mechanistic rules involving proliferation, cell death, and differentiation of endogenous and adoptively transferred NK cells in the expansion phase of the response to MCMV infection. We found that the interplay between cell proliferation and cell death of mature CD27- NK cells and a highly proliferative CD27-Ly6C- mature subtype and intrinsic stochastic fluctuations in these processes play key roles in regulating the heterogeneity and population of the NK cell subtypes. Furthermore, we estimate rates for maturation of endogenous NK cells in homeostasis and in MCMV infection and found that only NK cell growth rates, and not differentiation rates, are appreciably increased by MCMV. Taken together, these results quantify the differences between the kinetics of NK cell antigen-specific expansion from that of CD8+T cells and unique mechanisms that give rise to the observed heterogeneity in NK cell clones generated from single NK cells in the expansion phase.

Squidly harnesses enzyme functional hierarchy and contrastive learning to efficiently predict catalytic residues from sequence

amora@aithyra.ac.at (Ariane Mora) — Fri, 04 Sep 2026 00:00:00 +0000

Enzymes present a sustainable alternative to traditional chemical industries, drug synthesis, and bioremediation applications. Because catalytic residues are the key amino acids that drive enzyme function, their accurate prediction facilitates enzyme function prediction. Sequence similarity-based approaches such as BLAST are fast but require previously annotated homologues. Machine-learning (ML) approaches aim to overcome this limitation; however, current gold-standard ML-based methods require high-quality 3D structures limiting their application to large datasets. To address these challenges, we developed Squidly, a sequence-only tool that leverages contrastive representation learning with a biology-informed, rationally designed pairing scheme to distinguish catalytic from non-catalytic residues using per-token Protein Language Model embeddings. Squidly surpasses state-of-the-art ML annotation methods in catalytic residue prediction while remaining sufficiently fast to enable wide-scale screening of databases. We ensemble Squidly with BLAST to provide an efficient tool that annotates catalytic residues with high precision and recall for both in- and out-of-distribution sequences.

Individual differences drive social hierarchies in male mouse societies

jonathan.reinwald@zi-mannheim.de (Corentin Nelias) — Tue, 01 Sep 2026 00:00:00 +0000

Social hierarchies structure groups and confer advantages on high-ranking individuals. In mice, individual position in hierarchies may emerge situationally from current group compositions, or, alternatively, may remain largely stable across groups as an internalized feature. Dominance and subordination are expressed in behaviors like tube competitions or agonistic chasing. The interaction of these behaviors in the shaping of social position in larger male mouse groups remains largely unknown. To address these questions, we developed the NoSeMaze, a semi-naturalistic, open-source, modular platform that enables automated long-term tracking of unperturbed groups. Across more than 4000 mouse-days, hierarchies derived from incidental competitions in the integrated tube tests were non-despotic, transitive, and stable even when group compositions changed. This stability supports an internalized component of competition-based social rank. Chasing was also stable across contexts. Notably, chasing was concentrated among high-ranking individuals, consistent with ongoing negotiation of social rank among individuals at the upper end of the hierarchy. The link between chasing and social rank strengthened in groups with less well-defined rank structure, where mice rely more on aggressive signaling to assert their position. Chasing and social rank were associated with certain dimensions of simultaneously measured physical and cognitive features. In summary, high-dimensional tracking with the NoSeMaze reveals that social position in mice is multifaceted and shaped by stable dimensions of individual behavior that persist across changing social contexts. The approach thus enables longitudinal modeling of individuality and social position as key resilience factors.

Systematic analysis of network-driven adaptive resistance to CDK4/6 and oestrogen receptor inhibition using meta-dynamic network modelling

lan.nguyen@adelaide.edu.au (Anthony Hart) — Thu, 27 Aug 2026 00:00:00 +0000

Drug resistance inevitably emerges during the treatment of cancer by targeted therapy. Adaptive resistance is a major form of drug resistance, wherein the rewiring of protein signalling networks in response to drug perturbation allows drug-targeted protein activity to recover. This can occur in the continuous presence of the drug and enables cells to survive/grow. Simultaneously, molecular heterogeneity enables the selection of drug-resistant cancer clones that can survive an initial drug insult, proliferate, and eventually cause disease relapse. Despite their importance, the link between heterogeneity and adaptive resistance, specifically how heterogeneity influences protein signalling dynamics to drive adaptive resistance, remains poorly understood. Here, we have explored the relationship between heterogeneity, protein signalling dynamics, and adaptive resistance through the development of a novel modelling technique coined Meta Dynamic Network (MDN) modelling. We use MDN modelling to characterise how heterogeneity influences the drug-response signalling dynamics of the proteins that regulate early cell cycle progression and demonstrate that heterogeneity can robustly facilitate adaptive resistance associated dynamics for key cell cycle regulators. We determined the influence of heterogeneity at the level of both reaction coefficients and protein abundance and show that reaction coefficients are a much stronger driver of adaptive resistance. Owing to the mechanistic nature of the underpinning ordinary differential equation framework, we then identified a full spectrum of subnetworks capable of driving adaptive resistance dynamics in the key early cell cycle regulators. Finally, we show that single-cell dynamic data supports the validity of our MDN modelling technique and a comparison between our predicted resistance mechanisms and known CDK4/6 and oestrogen receptor inhibitor resistance mechanisms suggests MDN modelling can be deployed to robustly predict network-level resistance mechanisms for novel drugs and additional protein signalling networks.

Thymic selection of the T cell receptor repertoire is biased toward autoimmunity in females

david.klatzmann@sorbonne-universite.fr (Adrien Six) — Thu, 20 Aug 2026 00:00:00 +0000

Women represent about 80% of patients with autoimmune diseases. This may partly result from sex-based differences in T cell receptor (TCR) selection during thymocyte development, potentially influenced by hormones and the lower expression of the Autoimmune Regulator (AIRE) transcription factor in females. To investigate this, we analyzed sex-specific differences in TCR generation and selection. We examined TCR repertoires in double-positive thymocytes and single-positive thymic cells, including CD8⁺ and CD4⁺ effector T cells and regulatory T cells (Tregs), derived from male and female organ donors. Minimal sex-based differences were observed in V and J gene usage, and there were no notable differences in TCR repertoire diversity, complementarity-determining region 3 (CDR3) length, amino acid composition, or network structure. No TCR sequences were exclusive to either sex. However, female effector T cells exhibited a significantly higher prevalence of TCRs specific to self-antigens implicated in autoimmunity compared to males, while female Tregs showed a reduced frequency of such TCRs. These differences were not observed for TCRs targeting self-antigens unrelated to autoimmunity or antigens associated with cancer or viruses. Our findings identify a sex-specific imbalance in thymic selection of TCRs with autoimmunity-associated specificities, providing mechanistic insight into the increased susceptibility of women to autoimmune diseases.

Methylation clocks fail to generalize across genetically admixed individuals

tony@capralab.org (Anthony J Griswold) — Tue, 04 Aug 2026 00:00:00 +0000

Epigenetic aging clocks based on DNA methylation patterns across the genome have emerged as a potential biomarker for risk of age-related diseases, like Alzheimer’s disease (AD), and environmental and social stressors. However, methylation clocks have not been comprehensively validated in genetically diverse individuals. Here, we evaluate a set of first-, second-, and third-generation methylation clocks in 621 AD patients and matched controls from African American, Hispanic, and White cohorts. The clocks are less accurate at predicting age in genetically admixed cohorts compared to the White cohort, especially for those with substantial African ancestry. This decreased accuracy holds in >2500 individuals of European and African ancestry from three additional datasets. The clocks also fail to consistently identify age acceleration in admixed AD cases compared to controls. To explore potential causes for the lack of generalization of the clocks, we intersected clock CpGs with methylation, germline genetic variants, and methylation QTL (meQTL) data from global populations. We find differential methylation between African and European ancestry individuals is common for clock CpGs. Genetic variants rarely disrupt clock CpGs between populations, but a substantial fraction of clock CpGs have meQTL with significantly higher frequencies in African genetic ancestries. Our results demonstrate that methylation clocks often fail to predict age and AD risk when applied across populations and suggest avenues for improving their portability by considering differences in genetic and epigenetic patterns across human populations.

Benchmarking biochemical networks generated by large language models

jsaucerman@virginia.edu (B Adam Bates) — Fri, 31 Jul 2026 00:00:00 +0000

Computational models of biochemical networks provide frameworks for predicting how molecular cues guide cell decisions. These models are typically limited by the time-intensive manual curation required to extract network mechanisms from incomplete literature. Here, we test whether general-purpose large language models (LLMs) can generate accurate models of signaling and metabolic networks. We find that general-purpose LLMs generate 24–65% of the reactions of literature-curated signaling networks for cardiomyocyte hypertrophy, myofibroblast activation, and mechanosignaling. Further, logic-based models based on these networks predict responses to perturbations with accuracies of 6–33%. In the context of metabolic modeling, LLMs are able to generate 64–91% of the reactions within the core Escherichia coli metabolic network and demonstrate highly variable accuracies in predicting substrate utilization. Current general-purpose LLMs generate biochemical networks with moderate accuracy, and this study provides a pipeline and benchmarks to guide future improvements.

Dichotomy between extracellular signatures of active dendritic chemical synapses and gap junctions

rishi@iisc.ac.in (Richa Sirmaur) — Thu, 30 Jul 2026 00:00:00 +0000

Local field potentials (LFPs) are compound signals that represent the dynamic flow of information across the brain, which have been historically associated with chemical synaptic inputs. How do gap junctional inputs onto active compartments shape LFPs? We developed a methodology to record extracellular potentials associated with different patterns of gap junctional inputs onto conductance-based models. We found that synchronous inputs through chemical synapses yielded a negative deflection in proximal extracellular electrodes whereas those onto gap junctions manifested a positive deflection. Importantly, we observed extracellular dipoles only when inputs arrived through chemical synapses but not with gap junctions. Remarkably, hyperpolarization-activation cyclic nucleotide-gated channels, which typically conduct inward currents, mediated outward currents triggered by the fast voltage transition caused by synchronous inputs. With rhythmic inputs at different frequencies arriving through gap junctions, we found strong suppression of LFP power at higher frequencies as well as frequency-dependent differences in the spike phase associated with the LFP when compared to respective chemical synaptic counterparts. All observed differences in LFP were mediated by the relative dominance of synaptic currents vs. voltage-driven transmembrane currents with chemical synapses vs. gap junctions, respectively. Our analyses unveil a hitherto unknown role for active dendritic gap junctions in shaping extracellular potentials.

Active dendrites enable robust spiking computations despite timing jitter

tsjb2@cam.ac.uk (Michael E Rule) — Mon, 27 Jul 2026 00:00:00 +0000

Dendritic action potentials exhibit long plateaus of many tens of milliseconds, outliving axonal spikes by an order of magnitude. The computational role of these slow events seems at odds with the need to rapidly integrate and relay information throughout large nervous systems. We propose that the timescale of dendritic potentials allows for reliable integration of asynchronous inputs. We develop a physiologically grounded model in which the extended duration of dendritic spikes equips each dendrite with a resettable memory of incoming signals. This provides a tractable model for capturing dendritic nonlinearities observed in experiments and in more complex, detailed models. Using this model, we show that long-lived, nonlinear dendritic plateau potentials allow neurons to spike reliably when confronted with asynchronous input spikes. We demonstrate this model supports non-trivial computations in a network solving an association/discrimination task using sparse spiking that is subject to timing jitter. This demonstrates a computational role for the specific timecourse of dendritic potentials in situations where decisions occur quickly, reliably, and with a low number of spikes. Our results provide empirically testable hypotheses for the role of dendritic action potentials in cortical function, as well as a potential bio-inspired means of realising neuromorphic spiking computations in analog hardware.

Large-scale synthetic data enable digital twins of human excitable cells

ceclancy@ucdavis.edu (Colleen E Clancy) — Thu, 23 Jul 2026 00:00:00 +0000

Individual variability shapes how diseases manifest, how patients respond to therapy and how rare phenotypes arise. Conventional experimental approaches obscure variation by averaging which limits mechanistic insight and predictive accuracy. We present a computational framework that builds digital twins of human-induced pluripotent stem cell-derived cardiomyocytes from a single optimized voltage clamp experiment. The framework depends on massive synthetic datasets comprising simulated cells that span broad ionic and electrophysiological ranges. These synthetic data make it possible to control parameters precisely, explore biological variability comprehensively, and train models beyond the limits of experimental data. A neural network trained on synthetic data then inferred biophysical parameters from experimental recordings from live cells, reproducing distinct electrophysiological features. Our study unites computational modeling, data simulation, and learning to enable scalable, precise, individualized cardiac electrophysiology modeling and can be readily extended to any electrically active cell type.

Functional muscle networks as biomarkers of post-stroke motor impairment and therapeutic responsiveness

david.oreilly166@gmail.com (Andrea Turolla) — Thu, 23 Jul 2026 00:00:00 +0000

Standardised assessment of post-stroke motor impairment and treatment responsiveness remains a major clinical challenge. In this study, we tackle this challenge by applying a novel muscle network analysis framework to human stroke survivors undergoing intensive upper-limb motor training (O’Reilly & Delis, 2024). Our approach revealed distinct patterns of redundant and synergistic muscle interactions, collectively reflecting the diverse biomechanical roles of flexor- and extensor-driven networks. From these patterns, we derived new biomarkers that stratified patients by gross motor impairment severity and therapeutic responsiveness, each associated with unique physiological signatures. Remarkably, we identified a shift from redundancy to synergy in muscle coordination as a hallmark of effective motor recovery—a transformation supported by a more precise quantification of impairment over conventional approaches. These findings offer an in-depth characterisation of post-stroke motor recovery and establish a robust, independent tool for evaluating rehabilitation efficacy. Future research should employ this framework to identify biomarkers of activities- and participation-related functional recovery.

Quantitative computerized analysis demonstrates strongly compartmentalized tissue deformation patterns underlying mammalian heart tube formation

jorgendm@ujaen.es (Jorge N Domínguez) — Tue, 21 Jul 2026 00:00:00 +0000

The quantitative analysis of tissue deformation at cellular resolution remains an important challenge in mammalian organogenesis. Here, we developed a new computational workflow to extract regional and temporal patterns of tissue deformation, and applied it to a collection of live microscopy datasets from mouse cardiogenesis. We devised a method to track tissue deformation directly from time-lapse raw images and experimentally validated the method by comparison with actual cell tracks. We then used a machine-learning approach to temporally and spatially align different specimens and reconstruct a single statistical model of tissue motion, deducing maps of strain, anisotropy, and tissue growth. We also implemented a virtual fate mapping tool that allows tracking any initial position in the cardiac primordium onto the linear heart tube (HT). Our study reveals predominant local cellular coherence during the deformation of the cardiac tissue, whereas strong compartmentalization of tissue deformation patterns transforms the bilateral cardiac primordium into a 3D longitudinal HT. At the future outer curvature of the primitive tube, the ventricular chamber forms by expansion of the tissue in a hemi-barrel shape with two harnessing belts: one that constrains tissue expansion at the arterial pole and one that constrains the expansion at the venous pole. Our study provides a new approach to understanding heart morphogenesis and proposes a new model of primitive HT formation.

Experimental verification of the error minimization theory using non-standard genetic codes constructed in vitro

ichihashi@bio.c.u-tokyo.ac.jp (Norikazu Ichihashi) — Mon, 13 Jul 2026 00:00:00 +0000

All living systems use an almost identical standard genetic code (SGC), in which 20 amino acids are assigned non-randomly. According to the error minimization theory, amino acids are arranged to minimize the mutational effect on protein function, while experimental verification remains limited. Here, we constructed 10 non-standard genetic codes (non-SGCs) in vitro by reassigning three amino acids (Ala, Ser, and Leu) in vacant codons of the minimal genetic code consisting of 21 tRNAs. Most of these non-SGCs have a higher cost of amino acid replacement than the SGC, calculated based on three amino acid properties: polar requirement (PR), molecular volume (MV), and hydropathy index (HI). The protein function of three reporter genes expressed using these non-SGCs decreased similarly when random mutations were introduced into the genes, implying that the effect of mutations was similar across all the non-SGCs tested here. This result provides direct experimental evidence that mutational robustness does not significantly change in individual reporter protein activity within the range of mutational cost tested in this study (Cost_PR: 5.29–5.77, Cost_MV: 1848–2348, and Cost_HI: 3.27–5.10), which covers approximately 18.4% (PR), 37.6% (MV), and 50.8% (HI) of the possible cost range achievable among one million randomly-generated genetic codes.

Comparing the outputs of intramural and extramural grants funded by National Institutes of Health

bihutchins@wisc.edu (B Ian Hutchins) — Thu, 09 Jul 2026 00:00:00 +0000

Funding agencies use a variety of mechanisms to fund research. The National Institutes of Health in the United States, for example, employs scientists to perform research at its own laboratories (intramural research), and it also awards grants to pay for research at external institutions such as universities (extramural research). Here, using data from 1594 intramural grants and 97,054 extramural grants funded between 2009 and 2019, we compare the scholarly outputs from these two funding mechanisms in terms of number of publications, Relative Citation Ratio, and clinical metrics. We find that extramural awards are more cost-effective for producing outputs commonly used for academic evaluation, such as publications and citations (per dollar), while intramural awards are more cost-effective for generating research that influences future clinical work, more closely in line with the agency’s health goals. These findings provide evidence that institutional incentives associated with different funding mechanisms drive their comparative strengths.

Celldetective, an AI-enhanced image analysis tool for unraveling dynamic cell interactions

remy.torro@gmail.com (Beatriz Díaz-Bello) — Wed, 08 Jul 2026 00:00:00 +0000

Analysis of multimodal and multidimensional data capturing dynamic interactions between diverse cell populations is a current challenge in bioimaging, especially in the context of immunology and immunotherapy research. Here, we introduce Celldetective, an open-source Python-based software tool designed for high-performance end-to-end analysis of image-based in vitro immune and immunotherapy assays. Celldetective is purpose-built for multicondition, 2D multi-channel time-lapse microscopy of mixed cell populations. Although it is optimised for the needs of immunology assays, it is nevertheless broadly applicable to any biological system involving interacting cell populations. The software seamlessly integrates AI-based segmentation, tracking, and automated single-cell event detection, all within an intuitive graphical interface that supports interactive visualisation, annotation, and training options. We showcase its capabilities with original datasets of single immune effector cell interactions with an activating surface mediated by bispecific antibodies and pairwise interactions in antibody-dependent cell cytotoxicity events.

A coma pattern-based autofocusing method resolves bacterial cold shock response at single-cell level

shuqiang.huang@siat.ac.cn (Jinjuan Wang) — Mon, 06 Jul 2026 00:00:00 +0000

Imaging-based single-cell physiological profiling holds great potential for uncovering fundamental bacterial cold shock response (CSR) mechanisms, but its application is impeded by severe focus drift during rapid temperature downshifts required for CSR induction. Here, we introduce LUNA (Locking Under Nanoscale Accuracy), an innovative autofocusing method that leverages the coma pattern of detection light to characterize focus drift. LUNA improves the focusing precision down to 3 nm and extends the focusing range to at least 40 times the objective depth of focus. These advancements enable us to investigate the complete dynamics of bacterial single-cell CSR, revealing continuous cellular growth and division. We resolve a three-phase adaptation process characterized by distinct growth deceleration dynamics, and show that bacterial cells maintain robust size regulation and coordinate uniform adaptation to cold shock through synchronized growth and elapsed cycles. Notably, a model based on scattering theory reconciles the paradox between the growth lag of batch culture and continuous single-cell growth. These findings fundamentally transform our understanding of bacterial CSR and highlight LUNA’s excellent potential for expanding state-of-the-art research in biology.

Lipid packing contributes to the confinement of caveolae to the plasma membrane

richard.lundmark@umu.se (Aleksei Kabedev) — Mon, 06 Jul 2026 00:00:00 +0000

Lipid packing is a fundamental characteristic of bilayer membranes. Yet, we lack detailed mechanistic understanding of how lipid packing directly affects membrane-associated cellular processes. Here, we address this by focusing on caveolae, small Ω-shaped invaginations of the plasma membrane, which serve as key regulators of cellular lipid sorting and mechano-responses. In addition to caveolae coat proteins, the lipid membrane is a core component of caveolae that critically impacts their biogenesis, morphology, and stability. We show that the small compound Dyngo-4a adsorbs and inserts into the membrane, resulting in a dramatic dynamin-independent inhibition of caveola dynamics. Analysis of model membranes in combination with molecular dynamics simulations revealed that a substantial amount of Dyngo-4a was inserted and positioned at the level of cholesterol in the bilayer, affecting lipid order in a cholesterol-dependent manner. Dyngo-4a treatment resulted in decreased lipid packing of the plasma membrane. This prevented caveolae internalization and lateral diffusion without affecting their morphology, associated proteins, or the overall cell stiffness. Artificially increasing plasma membrane cholesterol levels was found to counteract the block in caveola dynamics caused by Dyngo-4a. Therefore, we propose that the outer leaflet lipid packing of cholesterol in the plasma membrane critically contributes to the confinement of caveolae to the plasma membrane.

SqueakPose Studio, an end-to-end platform for pose estimation and real-time edge-AI deployment

david.haggerty@nih.gov (Caleb Browning Darden) — Mon, 06 Jul 2026 00:00:00 +0000

Accurate pose estimation underpins quantitative analysis of behavior, yet many deep learning-based tracking tools remain optimized for offline workflows that rely on fragmented software pipelines, workstation-grade GPUs, or external middleware to enable real-time deployment. Here, we present an integrated software-hardware ecosystem for pose estimation that spans dataset creation, model training, offline analysis, and real-time deployment on embedded edge-computing devices. SqueakPose Studio provides a software suite for whole-frame, deep learning-based pose estimation that unifies dataset creation, manual and model-assisted labeling, model training, validation, and large-scale offline inference. The system leverages modern object-detection architectures to enable efficient end-to-end training and inference without patch-based sampling or multistage post-processing, and supports execution on CPUs, GPUs, and Apple Silicon. For experimental settings requiring continuous recording and synchronized data acquisition, SqueakView enables real-time model deployment, video capture, and sensor logging on embedded edge-computing hardware, while MouseHouse provides a compact, modular enclosure designed for home cage-based experiments that integrates embedded GPU compute, microcontroller-based timing, and peripheral I/O. A shared data format and deterministic timing architecture ensure consistency across offline analysis and real-time deployment. Together, SqueakPose Studio, SqueakView, and MouseHouse provide a unified platform for pose estimation that supports both conventional offline analysis and embedded, real-time experimentation, without reliance on workstation-grade hardware or external middleware.

TopoMetry systematically learns and evaluates the latent geometry of single-cell data

david.oliveira@dpag.ox.ac.uk (Ana I Domingos) — Fri, 03 Jul 2026 00:00:00 +0000

Reconstructing and investigating the geometry underlying data is a fundamental task in single-cell analysis, yet no unified framework exists for learning, evaluating, and diagnosing representations that faithfully preserve it. We present TopoMetry, a geometry-aware framework that learns intrinsic coordinate systems directly from the data and refines them into high-fidelity spectral scaffolds. These scaffolds capture both local neighborhoods and global structures, supporting downstream analyses such as clustering and visualization. In benchmarks across diverse single-cell datasets, TopoMetry preserved geometry more reliably than standard workflows and revealed biological signals otherwise obscured, including unexpected transcriptional diversity among T cells and links between RNA-defined subpopulations, and clonal expansion. The full analysis can be executed with a single line of code to generate a comprehensive report, making the framework both powerful and accessible. Beyond individual findings, TopoMetry warrants a shift of focus from static two-dimensional projections to the systematic learning and evaluation of geometry itself, enabling more accurate exploration of cellular diversity.

Generative modeling for RNA splicing prediction and design

yosephb@biociphers.org (Anna Tangiyan) — Fri, 29 May 2026 00:00:00 +0000

Alternative splicing (AS) of pre-mRNA plays a crucial role in tissue-specific gene regulation, with disease implications due to splicing defects. Predicting and manipulating AS can therefore uncover new regulatory mechanisms and aid in therapeutic design. We introduce TrASPr+BOS, a generative AI model with Bayesian Optimization for predicting and designing RNA for tissue-specific splicing outcomes. Transformer for Alternative Splicing Prediction (TrASPr) is a multi-transformer model that can handle different types of AS events and generalize to unseen cellular conditions. It then serves as an oracle, generating labeled data to train a Bayesian Optimization for Splicing (BOS) algorithm to design RNA for condition-specific splicing outcomes. We show TrASPr+BOS outperforms existing methods, enhancing tissue-specific AUPRC by up to 1.8-fold and capturing tissue-specific regulatory elements. We validate hundreds of predicted novel tissue-specific splicing variations and confirm new regulatory elements using dCas13. We envision TrASPr+BOS as a light yet accurate method researchers can probe or adopt for specific tasks.