XML / RSS feed — for news readers and subscription
Combination of dysfunctional beliefs about sleep and excessive daytime sleepiness as a psychobehavioral characteristic of comorbid insomnia and sleep apnea
Combination of dysfunctional beliefs about sleep and excessive daytime sleepiness as a psychobehavioral characteristic of comorbid insomnia and sleep apnea
Predicting protein cascade expression from H&E images
That author's affiliation: The Ohio State University
Institution (first & last author): The Ohio State University
by Alejandro Leyva, Abdul Rehman Akbar, Muhammad Khalid Khan Niazi
Protein expression within oncogenic or suppressive pathways is a hallmark indicator of oncogenesis. While traditional AI models in digital pathology attempt to predict singular proteins, there is a need to predict the downstream expression of proteins to indicate the propagation of signals. RNA expression provides novel information, but does not provide information about the downstream propagation of protein signals or whether those signals are functional. Using Reverse Phase Protein Array (RPPA) data with whole-slide images (WSIs) from the publicly available Cancer Genome Atlas Breast Adenocarcinoma dataset (TCGA-BRCA), we predict the expression of five key proteins identified from the apoptosis cascade, using DNA damage and repair (DDR) cascades as a biological control. Furthermore, we examine the performance of patch- level Vision Transformers (ViT) on the regression task, which was tested against the designed cellular-level ViT, CellRPPA. Our results demonstrate that patch-level vision transformers were unable to obtain statistically significant predictive results, achieving R-squared values < 0.1 for all folds. In addition, CellViT obtained R-squared values >0.1 in all five test folds. We also show that morphologically indicative cascades, such as the apoptosis cascade, provide significantly higher performance compared to the DDR cascade.Co-morbid biomarkers for sarcopenic obesity associated with gut microbiota metabolites: From burden to treatment
by Juehan Wang, Haijun Li, Weiyi Shi, Xiaoxu Ren, Yingying Liu, Lin Mao, Daming Wang, Tianfang Zhang, Ziwei Zhang, Huiqin Zheng, Xiaofeng Yang, Mingfei Yao, Zuobing Chen
BackgroundThe detrimental cycle of sarcopenic obesity (SO) significantly reduces quality of life in older adults, while the mechanisms are still unclear.
Materials and methodsWe first analyzed the incidence of SO using the CHARLS database. We identified key genes by integrating differentially expressed genes, weighted gene co-expression network analysis, and targets of gut microbiota metabolites, refining the selection through machine learning methods (LASSO, XGBoost, SVM-REF, Random Forest). These genes were validated through single-cell sequencing, receiver operating characteristic analysis, and Muscle immunohistochemistry in a high-fat-diet (HFD) induced mouse model. Further analyses comprised immune infiltration profiling, pathway enrichment, and transcriptional regulation analysis. Additionally, we explored the relationships between key genes and autophagy, ferroptosis, and immunity responses. Finally, we predicted and evaluated potential therapeutic compounds via the CMap database and molecular docking.
ResultsSO incidence in China increased significantly from 16.1% (2011) to 20.4% (2018). Machine learning identified ALDH1A3, CSF1R, and PHGDH as key genes. These genes were validated in external muscle single-cell datasets, demonstrating robust diagnostic performance with AUC values exceeding 0.72 across four independent GEO cohorts. Following an HFD intervention in mice, ALDH1A3 and CSF1R expression in muscle tissue was significantly upregulated, while PHGDH showed a consistent upward trend that did not reach statistical significance. Immune infiltration analysis revealed a significant increase in resting NK cells in both obesity and sarcopenia states. Functional enrichment analyses using Gene Ontology and Kyoto Encyclopedia of Genes and Genomes linked the genes to transcriptional regulation pathways. The Cisbp_M4923 motif was identified as the most relevant transcription factor binding site. Finally, molecular docking simulations indicated stable binding of the top candidate compound, Birinapant, to the key gene targets.
ConclusionALDH1A3, CSF1R, and PHGDH serve as potential co-morbid biomarkers for SO.
VUStruct: A compute pipeline for high throughput and personalized structural biology
That author's affiliation: Vanderbilt University
Institution (first & last author): Vanderbilt University
by Christopher W. Moth, Jonathan H. Sheehan, Abdullah Al Mamun, R. Michael Sivley, Alican Gulsevin, David C. Rinker, Zenab F. Mchaourab, Undiagnosed Diseases Network , John A. Capra, Jens Meiler
Effective diagnosis and treatment of rare genetic disorders requires the interpretation of a patient’s genetic variants of unknown significance (VUSs). Today, clinical decision-making is primarily guided by gene-phenotype association databases and DNA-based scoring methods. Our web-accessible variant analysis pipeline, VUStruct, supplements these established approaches by deeply analyzing the downstream molecular impact of variation in context of 3D protein structure. VUStruct’s growing impact is fueled by the co-proliferation of protein 3D structural models, gene sequencing, compute power, and artificial intelligence. Contextualizing VUSs in protein 3D structural models also illuminates longitudinal genomics studies and biochemical bench research focused on VUS, and we created VUStruct for clinicians and researchers alike. We now introduce VUStruct to the broad scientific community as a mature, web-facing, extensible, High-Performance Computing (HPC) software pipeline. VUStruct maps missense variants onto automatically selected protein structures and launches a broad range of analyses. These include energy-based assessments of protein folding and stability, pathogenicity prediction through spatial clustering analysis, and machine learning (ML) predictors of binding surface disruptions and nearby post-translational modification sites. The pipeline also considers the entire input set of VUS and identifies genes potentially involved in digenic disease. VUStruct’s utility in clinical rare disease genome interpretation has been demonstrated through its analysis of over 175 Undiagnosed Disease Network (UDN) Patient cases. VUStruct-leveraged hypotheses have often informed clinicians in their consideration of additional patient testing, and we report here details from two cases where VUStruct was key to their solution. We also note successes with academic research collaborators, for whom VUStruct has informed research directions in both computational genomics and wet lab studies.Explainable AI-driven diagnosis model for early glaucoma detection using grey-wolf optimized extreme learning machine approach
That author's affiliation: Harvard University
First author institution: Dr. C. V. Raman University
Last author institution: Harvard University
by Debendra Muduli, Santosh Kumar Sharma, Sujata Dash, Bernardo Lemos, Saurav Mallik
Glaucoma is a leading global cause of blindness, making early detection essential. This paper introduces GlaucoXAI (Glaucoma Explainable AI), an advanced computer-aided diagnosis (CAD) model that integrates machine learning and explainable AI for glaucoma detection using retinal fundus images. The proposed model consists of four stages, including preprocessing, feature extraction, dimensionality reduction, and classification. Initially, features are extracted using the fast discrete curvelet transform with wrapping (FDCT-WRP) to obtain curve-type features. During the next stage, principal component analysis (PCA) and linear discriminant analysis (LDA) are combined to reduce the dimensionality of the feature matrix, followed by a classification stage employing an improved grey wolf optimization (IMGWO) with an extreme learning machine (ELM) to optimize the weight and bias to reduce the overfitting of the model. The model has been experimented with two publicly available datasets named G1020 and ORIGA. The model has achieved 93.87% accuracy on G1020 and 95.38% on ORIGA, outperforming existing methods. The 10 × 5-fold stratified cross-validation (SCV) with explainable AI enhances the interpretability of models and improves clinician trust. Overall, the proposed approach offers accurate, efficient, and explainable glaucoma diagnosis, potentially supporting ophthalmologists in early disease detection.NAT10 maintains stem cell homeostasis by mitigating mRNA decay through an ac<sup>4</sup>C-independent mechanism
Li, Huo, Zhang, Liu and colleagues report that the transcripts of genes involved in haematopoietic stem cell function share a protective mechanism against RNA decay through a non-canonical function of NAT10, independent of its N4-acetylcytidine catalytic activity.
SARS-CoV-2 spike protein exerts an anti-cancer effect in A549 cells in association with MEG3 and BCYRN1 regulation
SARS-CoV-2 spike protein exerts an anti-cancer effect in A549 cells in association with MEG3 and BCYRN1 regulation
Learning the bistable cortical dynamics of the sleep-onset period
That author's affiliation: University of Washington
Institution (first & last author): Centre National de la Recherche Scientifique
by Zhenxing Hu, Manaoj Aravind, Xu Lei, J. Nathan Kutz, Jean-Julien Aucouturier
Humans just don’t fall asleep like a log – or step-function. Rather, the sleep-onset period (SOP) exhibits dynamic and non-monotonous changes of electroencephalogram (EEG) with high, and so far poorly understood, intra- and inter-individual variability. Computational models of the sleep regulation network have suggested that the transition to sleep can be viewed as a noisy bifurcation at a saddle node which is determined by an underlying control signal or “sleep drive”. However, such models do not describe how internal control signals in the SOP can produce rapid switches between stable wake and sleep states, nor how these state-space changes are translated in the macroscopic EEG. Here, we propose a minimally-parameterized stochastic dynamical model, in which one slowly-varying control parameter drives the wake-to-sleep transition while exhibiting noise-driven bistability. We provide a procedure for estimating the parameters of the model given single observations of experimental sleep EEG data, and show that it can reproduce a wide variety of SOP phenomenology. Using the model to analyze a pre-existing sleep EEG dataset, we find that the estimated model parameters correlate with subjective sleepiness reports. These results suggest that the bistable characteristics of the SOP can serve as biomarkers for tracking intra- and inter-individual variability of sleep-onset disorders.Evolutionary Kuramoto dynamics unravels origins of chimera states in neural populations
That author's affiliation: City St George's, University of London
Institution (first & last author): Dartmouth College
by Thomas Zdyrski, Scott Pauls, Feng Fu
Neural synchronization is central to cognition. However, incomplete synchronization often produces chimera states, where coherent and incoherent dynamics coexist. Recent studies have suggested that these chimera states could be important in human cognitive organization. In particular, chimera states have been suggested as a regulator of cognitive integration and regulation with varying quality as humans age. While previous studies have explored such chimera states using networks of coupled oscillators, it remains unclear why neurons commit to communication or how chimera states persist. Here, we investigate the coevolution of neuronal phases and communication strategies on directed, weighted networks where interaction payoffs depend on phase alignment and may be asymmetric due to unilateral communication. The graph structure enables us to apply a game-theoretic model of Kuramoto-like oscillators to brain connectomes, and the asymmetry captures biochemical differences between communicative and non-communicative neurons. Combined, these two generalizations enable us to apply the computationally-tractable game-theoretic model of Kuramoto models to realistic brain networks and analyze the role of connectome structure on neuron communication. We find that both connection weights and directionality influence the stability of communicative strategies—and, consequently, full synchronization—as well as the strategic nature of neuronal interactions. Applying our framework to the C. elegans connectome, we show that emergent payoff structures, such as the staghunt game, control population dynamics. We demonstrate that weighted, directed connectivity in the Caenorhabditis elegans (C. elegans) connectome is sufficient to generate robust chimera states modulated by payoff asymmetries. Our computational results demonstrate a promising neurogame-theoretic perspective, leveraging evolutionary graph theory to shed light on mechanisms of neuronal coordination beyond classical synchronization models.Evaluating the utility of amino acid similarity-aware kmers to represent TCR repertoires for classification
That author's affiliation: University of Liverpool
Institution (first & last author): University of Liverpool
by Hannah Kockelbergh, Shelley C. Evans, Liam Brierley, Peter L. Green, Andrea L. Jorgensen, Elizabeth J. Soilleux, Anna Fowler
Insights gained through interpretation of models trained on the T-cell receptor (TCR) repertoire contribute to advances in understanding of immune-mediated disease. This has the potential to improve diagnostic tests and treatments, particularly for autoimmune diseases. However, TCR repertoire datasets with samples from donors of known autoimmune disease status generally include orders of magnitude fewer samples than TCR sequences. Promising TCR repertoire classification approaches consider relationships between non-identical TCR sequences. In particular, kmer methods demonstrate strong and stable performance for small datasets. We propose a TCR repertoire representation that considers the relationships between amino acids within kmers flexibly and efficiently. XGBoost and logistic regression models are trained and tested on kmer representations of TCR repertoire datasets including samples from patients with coeliac disease as well as donors with previous cytomegalovirus infection. XGBoost models outperform logistic regression, indicating that interactions may be crucial for discriminative ability. We find that a reduced alphabet based on BLOSUM62 can lead to a model with slightly stronger XGBoost testing performance than other kmer features. Though it remains unclear whether there is an amino acid encoding that can substantially improve TCR repertoire classification with reduced alphabet kmers, evidence that this representation enables faster training of XGBoost models in comparison to kmer clusters suggests that our reduced alphabet approach permits wider exploration of amino acid similarity in practice. Finally, we detail motifs which are important in each top-performing XGBoost model and compare them to TCR sequences previously associated with each immune status. We highlight the challenge of interpreting non-linear TCR repertoire classification models trained on kmers which, if overcome, could lead to biomarker discovery for autoimmune diseases.Complexity of resting cortical activity predicts neurophysiological responses to theta-burst stimulation but fails to generalize: A rigorous machine-learning approach
That author's affiliation: Harvard Medical School, Beth Israel Deaconess Medical Center
First author institution: University of Texas MD Anderson Cancer Center
Last author institution: Harvard Medical School, Beth Israel Deaconess Medical Center
by Matthew Herbert Ning, Haoqi Sun, Brice Passera, Duygu Bagci Das, Brandon Westover, Alvaro Pascual-Leone, Emiliano Santarnecchi, Mouhsin M. Shafi, Recep A. Ozdemir
BackgroundSubstantial variability in individual responses to intermittent theta-burst stimulation (iTBS) limits its clinical efficacy, yet neurophysiological mechanisms underlying this variability remain unclear. While most machine-learning studies have focused on modeling behavioral or clinical effects of repetitive transcranial magnetic stimulation (rTMS), the few studies examining neurophysiological outcomes utilized limited feature sets in single-visit settings, which captured only inter-subject variability and most importantly lacked independent validation sets.
MethodsTo address these gaps, we employed supervised machine learning models that integrated baseline resting-state EEG (rsEEG) features and baseline transcranial magnetic stimulation (TMS)-evoked measures, including motor-evoked potentials (MEPs) and TMS-evoked potentials (TEPs), to predict neurophysiological responses to a single iTBS session applied over the primary motor cortex in two independent test-retest studies of healthy adults. We also employed statistical and reliability analysis to understand the statistical relationship between resting state EEG and responses to iTBS.
ResultsInternal cross-validation within the training cohort yielded promising binary classification performance (accuracy: 81%), identifying coarse-grained multiscale distribution entropy of rsEEG as the most robust predictor of local cortical excitability changes indexed by the 100–131ms window of TEPs. However, predictive performance markedly declined upon external validation (accuracy: 69%), reflecting unstable relationships between predictors and outcomes likely driven by substantial intra- and inter-individual variability of iTBS-induced changes in neurophysiological outcomes.
ConclusionsThese findings emphasize that while EEG complexity measures can capture baseline brain states relevant for neuromodulation to a certain degree, the inherent instability of single-session iTBS effects significantly constrains model generalizability and underscores the necessity of test-retest paradigm to avoid overly optimistic performance estimates. Future studies with multi-session and individualized stimulation protocols are urgently needed to better characterize neurophysiological mechanisms underlying rTMS effects and ultimately enhance its therapeutic potential.
Characterizing the metabolomes of microglia, astrocytes and neurons in ageing and Alzheimer’s brains
That author's affiliation: University of Texas Southwestern Medical Center First author institution: Shanghai Jiao Tong University Last author institution: University of Texas Southwestern Medical Center
Yu, Li and colleagues explore the metabolome of neurons, microglia and astrocytes under normal, ageing and Alzheimer’s disease conditions. They find enrichment of glutathione and polyamine metabolism in microglia, which decreased with ageing and in Alzheimer’s disease.
CASER: A semi-supervised model with multi-omics data integration prioritizes cancer-associated epigenetic regulator genes
That author's affiliation: Salk Institution for Biological Studies
First author institution: Unknown
Last author institution: Wenzhou Medical University
by Hao Li, Chaohuan Lin, Liyu Liu, Jie Lyu, Zhen Feng
Prioritizing a reliable list of cancer-associated epigenetic regulators (cERs) is critical for cancer diagnosis and discovery of drug targets. While various cERs have been proposed to play important roles as cancer drivers, we anticipate that further cERs can be identified through computational analyses. In this study, we introduce a semi-supervised machine-learning approach based on tri-training model, termed Cancer-ASsociated Epigenetic Regulator identification (CASER). CASER integrates a wide range of multi-omics-derived features, including mutational, genomic, epigenetic, and transcriptomic data, to prioritize cERs as well as the four functional subtypes of cERs. When evaluated against an independent gene set, CASER demonstrates superior predictive performance compared to various other supervised machine-learning and deep semi-supervised models. CASER identified novel cERs that demonstrated cancer-driving potential and essentiality for cell survival. These novel cERs were comparable to established cancer driver genes and outperformed existing approaches for cER prediction. CASER identified dozens of novel cERs, of which six candidate cERs were shown to have roles in altering cell proliferation in four cancer cell lines. Furthermore, the prioritized cERs, particularly dual-role cERs, are more associated with anti-cancer medicines, underscoring their potential as therapeutic targets in cancer. Our study can offer valuable insights of cERs for future functional studies, advancing the understanding of their role in cancer biology.Bifurcation of neural firing patterns driven by potassium dynamics and neuron–electrode geometry during high-frequency stimulation
That author's affiliation: Johns Hopkins School of Medicine
First author institution: Johns Hopkins School of Medicine
Last author institution: Unknown
by Yue Yuan, Junyang Zhang, Chen Wang, Hao Yan, Ning Zhang, Kun Zhang, Zheshan Guo, Zhaoxiang Wang
High-frequency stimulation (HFS), the basis of deep brain stimulation, elicits diverse neuronal responses, yet the mechanisms remain unclear. Classical conduction block theories cite sodium channel inactivation and axonal failure but cannot explain the abrupt, reproducible firing transitions observed in vivo. Here, we combine single-unit recordings from rat CA1 neurons with a biophysically detailed multi-compartment model to examine how HFS shapes axonal excitability. The results show that neuronal responses are governed by two coupled factors: the electrode–axon geometry and peri-axonal extracellular potassium ([K⁺]o) dynamics. Small changes in either parameter reliably triggered bifurcation-like transitions between tonic, clustered, and low-rate regular firing. Conduction block preceded initiation failure with increasing electrode-axon distance, whereas elevated [K⁺]o shifted membranes between excitable and non-excitable states. This unified bifurcation framework extends the conduction block hypothesis, recasts axons as nonlinear elements, and provides mechanistic insights to optimize electrode placement, stimulation tuning, and closed-loop neuromodulation strategies.Cancer-associated fibroblasts regulate DNA repair in pancreatic cancer through NDRG1-mediated R-loop processing
That author's affiliation: University of Bergen Institution (first & last author): Beth Israel Deaconess Medical Center
Kozlova et al. report a functional role for NDRG1 in regulating R-loops and DNA repair, thereby linking cancer-associated fibroblast-derived extracellular matrix to chemotherapy responses in pancreatic cancer.
Nine quick tips for software containerization
That author's affiliation: University of Auckland
Institution (first & last author): University of Auckland
by David Moreau, Kristina Wiebels
Software containerization has become a cornerstone of modern computational biology, enabling researchers to package code, dependencies, and execution environments in portable and reusable units. Containers support reproducibility, facilitate collaboration, and lower barriers to deploying complex computational workflows across heterogeneous systems. At the same time, inappropriate or superficial use of containers can undermine these benefits, leading to brittle environments, security risks, or false confidence in reproducibility. In this article, we present nine practical and actionable tips for using software containers effectively in computational biology research. Rather than focusing narrowly on container syntax or tooling, we address conceptual decisions that arise throughout the research lifecycle: when containerization is appropriate, how to balance reproducibility with flexibility, how to manage dependencies and data, and how to share containers responsibly. These tips are intended for researchers with varying levels of experience, from those adopting containers for the first time to those maintaining mature, containerized workflows.Automating population construction and parallel simulation of biophysical models for neuromuscular cells: An inverse approach
That author's affiliation: Korea Institute of Science and Technology
Institution (first & last author): Korea Institute of Science and Technology
by Hojeong Kim
Biophysical modeling and simulation help to promote a comprehensive understanding of the neuromuscular mechanisms underlying muscle force generation and control in normal and pathological states. However, this process is labor intensive and limited to special conditions due to the heterogeneity of neuromuscular cells and the variability in their organization across body parts and ages. We present a methodology to resolve this issue. First, we formulate a building-block approach with an inverse modeling framework for automated population construction and tractable hierarchical analysis under various physiological conditions. Second, we devise a network folder-based approach with a virtual environment technique for efficient parallel simulation that can operate on a multicore computer, a supercomputing system, or a computer network through the internet. Third, we implement the methodology by developing open-source command-line software called pNMS. Finally, we demonstrate that pNMS can replicate experimental and simulation results from different environments and predict the population behaviors of neuromuscular cells depending on their organization and muscle length. With an intuitive, flexible application programming interface, this software tool may offer a solution for promoting efficient investigation and an in-depth understanding of neuromuscular function at cellular resolution under realistic scenarios.Structural mechanisms underlying distinct binding and activities of 18:0 and 18:1 lysophosphatidic acids at LPA1 receptor
That author's affiliation: Washington State University
Institution (first & last author): Washington State University
by Ayobami Diyaolu, Peter Obi, Pravita Balijepalli, Kathryn E. Meier, Senthil Natesan
Lysophosphatidic acids (LPAs) are bioactive lipids that regulate numerous physiological functions in humans. Cell signaling by LPAs is mediated mainly via six LPA receptors (LPA1-6), class A G protein-coupled receptors (GPCRs). Among these, LPA1 is recognized to play an essential role in cell proliferation, survival, migration, and tumorigenesis. Despite the structural similarity, 18:0-LPA and 18:1-LPA exhibit distinct functional responses in cell lines overexpressing LPA1. Specifically, our in vitro studies show that 18:1-LPA induces greater Erk activation than 18:0-LPA in PC-3 human prostate cancer cells. The structural basis underlying this differential receptor activation has not been previously studied. Using classical molecular dynamics and enhanced sampling techniques, we examined the access and binding mechanisms of the two LPA species to the active state LPA1 receptor. The results show that 18:0-LPA and 18:1-LPA adopt distinct and dynamic poses in the orthosteric pocket despite their similar starting configurations. Mainly, the alkyl chains of the ligands exhibit distinct orientations and residue interactions, leading to differential conformational changes in key activation switches on the conserved CWxP and PIF structural motifs of the receptor. Also, there are significant differences in interhelical interactions at the intracellular end of the transmembrane helices 1, 3, 6, and 7. These distinct arrangements lead to striking differences in LPA1 interactions with the Gα-helix of the heterotrimeric Gi-protein. Notably, 18:0-LPA and 18:1-LPA exhibit similar membrane partitioning characteristics and receptor entry processes through aqueous paths. Our comprehensive in-silico studies offer valuable structural insights into the observed differences in functional responses by 18:0- and 18:1-LPA.Electrophilic compound screening identifies GPX4-dependent ferroptosis as a senescence vulnerability
D’Ambrosio et al. screen 10,480 electrophilic compounds for senolytic effects and show that senolytic chloroacetamides or GPX4 inhibitors selectively kill senescent cells by ferroptosis.
Using experimental results of protein design to guide biomolecular energy-function development
by Hugh K. Haddox, Gabriel J. Rocklin, Francis C. Motta, Devin Strickland, Samer F. Halabiya, Cameron Cordray, Hahnbeom Park, Eric Klavins, David Baker, Frank DiMaio
Computational models of macromolecules have many applications in biochemistry, but physical inaccuracies limit their utility. One class of models uses energy functions rooted in classical mechanics. The standard datasets used to train these models are limited in diversity, pointing to a need for new training data. Here, we sought to explore a new paradigm for training an energy function, where the Rosetta energy function was used to design de novo proteins. Experimental results on these designs were then used to identify failure modes of design, which were subsequently used as a “guiding principle” to retrain the energy function. Specifically, we examined a diverse set of de novo protein designs experimentally tested for their ability to stably fold, identifying unstable designs that were predicted to be stable by the Rosetta energy function. Using deep mutational scanning, we identified single amino-acid mutations that rescued the stability of these designs, providing insight into common failure modes of the energy function. We identified one key failure mode, involving steric clashing in protein cores. We identified similar overpacking when using Rosetta to refine high-resolution protein crystal structures, quantified the degree of overpacking, and refit a small set of energy-function parameters to better recapitulate native-like packing. Following fitting, we largely eliminated the failure mode in the refinement task, while retaining performance on other benchmarks, resulting in an updated version of the Rosetta energy function. This work shows how learning from protein designs can guide energy-function development.Spatio–temporal modelling of <i>in vitro</i> influenza A virus infection: The impact of defective interfering particles on the type I interferon response
by Yimei Li, Bjarke Frost Nielsen, Simon A. Levin, Aartjan J.W. te Velthuis, Bryan T. Grenfell
Defective interfering particles (DIPs) are incomplete viral genomes that modulate infection by competing with wild–type viruses and activating the innate immune response. Activation of the immune response leads to the production of cytokines and chemokines, including type I interferon (IFN), which restricts viral growth and may cause cell death. How DIPs interact with type I interferon (IFN) in spatially structured environments remains unclear. Focusing here on influenza A viruses, we developed a spatially explicit, stochastic model of in vitro viral infection that integrates virus and DIP replication, IFN signalling, and alternative dispersal modes. We find that: (1) our model captures the ring–like and patchy plaque morphologies observed experimentally; (2) IFN production peaks at an intermediate DIP ratio, reflecting a trade–off between early immune activation and sufficient co–infection; and (3) even a small fraction of long–range spread by virus and DIPs enables escape from the immune-based containment despite long-range IFN diffusion; this causes stronger antiviral responses but earlier peaks in virus egress at similar levels of cell loss. The model is available as an interactive platform: https://shiny-spatial-infection-app-production.up.railway.app/.DMAPLM: A multimodal pretrained framework for computational drug repositioning
by Hailin Chen, Zhongling Li
Drug repositioning offers an efficient route to discover new therapeutic indications for existing drugs. However, current computational drug repositioning models often face challenges related to data scarcity, heterogeneity, and therefore limited generalizability. To address these limitations, this study introduces DMAPLM, a multimodal pretrained framework for predicting drug-disease associations for further drug repositioning screening. DMAPLM leverages a lightweight dual-encoder architecture, utilizing ChemBERTa-2 for molecular encoding of drug SMILES strings and BioBERT for semantic encoding of multi-field disease texts. The framework explicitly aligns drug and disease representations through contrastive learning and employs attention-weighted pooling to emphasize informative molecular substructures. A Random Forest classifier is finally used for association prediction based on the enhanced multimodal features. We compile a new and comprehensive benchmark dataset for performance evaluation. Extensive experiments demonstrate that DMAPLM significantly outperforms six state-of-the-art baseline models, achieving an AUROC of 0.8919 and AUPR of 0.9116 under five-fold cross-validation, representing an improvement of up to 9%. Furthermore, DMAPLM exhibits robust performance in challenging cold-start scenarios, highlighting its practical utility for identifying novel drug-disease relationships. Case studies along with molecular docking analysis confirm the interpretability and biological meaningfulness of our predictions. Our study provides a powerful and interpretable approach for computational drug repositioning.Sharing the spotlight: Uncovering common attentional dynamics across species
That author's affiliation: Society for Neuroscience
Institution (first & last author): Society for Neuroscience
by Mina Glukhova, Alejandro Tlaie, Robert Taylor, Pierre-Antoine Ferracci, Katharine Shapcott, Berkutay Mert, Olga Arne, Andrei Ciuparu, Raul C. Muresan, Martha N. Havenith, Marieke L. Schölvinck
Sustained attention is a key underlying process to many natural behaviours that are shared across species. Yet the way attention is commonly studied in a lab context precludes meaningful cross-species comparisons. Here, we engaged mice, monkeys, and humans in the same, naturalistic perceptual decision task in a virtual reality environment. We captured their behaviour in several parameters along the speed/accuracy axes along which sustained attention is classically defined, and used Hidden Markov Models (HMMs) to infer four attentional states. We show that the dynamics of these states, both in terms of their durations and transitions, are more similar across species than might have been expected. Moreover, attentional state fluctuations seem to be internally generated and are not predicted by task attributes. The task and analyses developed here represent a new approach to infer the dynamics of sustained attention from naturalistic behaviours, in a way that is generalizable across species.Phase-field modeling of border cell cluster migration in <i>Drosophila</i>
by Naghmeh Akhavan, Alexander George, Michelle Starz-Gaiano, Bradford E. Peercy
Collective cell migration is a fundamental biological process that drives events such as embryonic development, wound healing, and cancer metastasis. In this study, we develop a biophysically informed phase-field model to investigate the collective migration of the border cell cluster in the Drosophila melanogaster egg chamber. Our model captures key aspects of the egg chamber architecture, including the oocyte, nurse cells, and surrounding epithelium, and incorporates both mechanical forces and biochemical cues that guide cell migration. We introduce the Tangential Interface Migration (TIM) force which captures contact-mediated propulsion generated along interfaces between the border cell cluster and surrounding nurse cells. Our simulations reveal three key features of TIM-driven migration that distinguish it from previous forms of chemotaxis: (1) the explicit nature of border cell–nurse cell overlap to initiate movement (i.e., border cells cannot move without a nurse cell substrate), (2) motion is tangential to border cell-nurse cell interfaces, and (3) persistent migration occurs even in regions where the spatial slope of chemoattractant is decreasing. Additionally, we demonstrate that with or without geometry-mediated alterations in chemoattractant distribution such as at intercellular junctions, we can vary induced migration pauses, independent of mechanical confinement. We capture an experimentally observed transition to dorsal migration at the oocyte with a sustained medio-lateral chemical cue of small amplitude. The results show how spatial constraints and interfacial forces shape collective cell movement and highlight the utility of phase-field models in capturing the interplay between tissue geometry, contact forces, and chemical signaling.Enhancing generalizability of model discovery across parameter space with multi-experiment equation learning for biological systems
by Maria-Veronica Ciocanel, John T. Nardini, Kevin B. Flores, Erica M. Rutter, Suzanne S. Sindi, Alexandria Volkening
Agent-based modeling (ABM) is a powerful tool for understanding self-organizing biological systems, but it is computationally intensive and often not analytically tractable. Equation learning (EQL) methods can derive continuum models from ABM data, but they typically require extensive simulations for each parameter set, raising concerns about generalizability. In this work, we extend EQL to Multi-experiment equation learning (ME-EQL) by introducing two methods: (i) one-at-a-time ME-EQL (OAT ME-EQL), which learns individual models for each parameter set and connects them via interpolation, and (ii) embedded structure ME-EQL (ES ME-EQL), which builds a unified model library across parameters. We demonstrate these methods by learning continuum models from a noisy birth–death mean-field model and from an on-lattice agent-based model of birth, death, and migration with spatial structure, often used to investigate cell biology experiments. We show that both methods significantly reduce the relative error in recovering parameters from agent-based simulations, with OAT ME-EQL offering better generalizability across parameter space. Our findings highlight the potential of equation learning from multiple experiments to enhance the generalizability and interpretability of learned models for complex biological systems.Monotherapy cancer drug-blind response prediction is limited to intraclass generalization
That author's affiliation: 3Argonne National Laboratory, Lamont, IL
Institution (first & last author): Mayo Clinic in Arizona
by William G. Herbert, Nicholas Chia, Paul A. Jensen, Marina R. S. Walther-Antonio
Monotherapy cancer drug response prediction (DRP) models predict the response of a cell line to a given drug. Analyzing these models’ performance includes assessing their ability to predict the response of cell lines to new drugs, i.e., drugs that are not in the training set. Drug-blind prediction displays greatly diminished performance or outright failure across a wide range of model architectures and different large pharmacogenomic datasets. Drug-blind failure is hypothesized to be caused by the relatively limited set of drugs present in these datasets. The time and cost associated with further cell line experiments is significant, and it is impossible to predict beforehand how much data would be enough to overcome drug-blind failure. We must first define how current data contributes to drug-blind failure before attempting to remedy drug-blind failure with further data collection. In this work, we quantify the extent to which drug-blind generalizability relies on mechanistic overlap of drugs between training and testing splits. We first identify that the majority of mixed set DRP model performance can be attributed to drug overfitting, likely inhibiting generalization and preventing accurate analysis. Then, by specifically probing the drug-blind ability of models, we reveal the sources of generalizable drug features are confined to shared mechanisms of action and related pathways. Furthermore, we observed that, for certain mechanisms, we can significantly improve performance by limiting the training of models to a single mechanism compared to training on all drugs simultaneously. Across multiple different model architectures examined in this paper, we observe that drug-blind performance is a poor benchmark for DRP as it does not describe model behavior, it describes dataset behavior. Our investigation displays that these deep learning models trained on large, monotherapy cell line panels can more accurately describe mechanism of action of drugs rather than their advertised connection to broader cancer biology.La benchmarking large language models for extracting biobank-derived insights into health and disease
by Manuel Corpas, Alfredo Iacoangeli
Biobank-scale datasets such as the UK Biobank have become foundational resources for advancing biomedical discovery. Yet the complexity and heterogeneity of these resources, spanning genomics, imaging, clinical records, and metadata, pose substantial barriers to access and interpretation. Large Language Models (LLMs) offer a promising avenue for making such datasets more navigable through natural language interfaces. However, the extent to which current general-purpose LLMs can retrieve and synthesize biobank-specific insights has not yet been systematically evaluated. In this study, we present a reproducible, multi-metric evaluation framework to benchmark the capabilities of leading LLMs. We evaluated six leading large language models: Gemini 3 Pro, Claude Opus 4.5, Claude Sonnet 4, GPT-5.2, Mistral Large, and DeepSeek V3, on four benchmark tasks designed to assess biobank-related knowledge retrieval. We evaluate model performance across six dimensions (coverage, semantic accuracy, factual correctness, domain knowledge, reasoning quality, and biobank specificity) and assessed output consistency using curated UK Biobank references and a robust random baseline. All models outperformed the baseline by 16× to 25 × , with strong statistical separation (p < 0.001), confirming meaningful biobank-specific knowledge retrieval. Gemini 3 Pro achieved the highest overall accuracy across tasks such as keyword synthesis, institution recognition, and topic inference, while Claude Sonnet 4 demonstrated the most uniform performance across evaluation dimensions. Our benchmark provides a rigorous framework for evaluating LLMs in biomedical settings. Using the UK Biobank as a real-world testbed, we highlight both the capabilities and limitations of current models, measuring their capacity to recall structured biomedical knowledge consistent with authoritative biobank metadata.‘Backpropagation and the brain’ realized in cortical error neuron microcircuits
by Kevin Max, Ismael Jaras, Arno Granier, Katharina A. Wilmes, Mihai A. Petrovici
Neural responses to mismatches between expected and actual stimuli have been widely reported across different species. How does the brain use such error signals for learning? While global error signals can be useful, their ability to learn complex computation at the scale observed in the brain is lacking. In comparison, more local, neuron-specific error signals enable superior performance, but their computation and propagation remain unclear. Motivated by the breakthrough of deep learning, this has inspired the ‘backpropagation and the brain’ hypothesis, i.e., that the brain implements a form of the error backpropagation algorithm. In this work, we introduce a biologically motivated, multi-area cortical microcircuit model, implementing error backpropagation under consideration of recent physiological evidence. We model populations of cortical pyramidal cells acting as representation and error neurons, with bio-plausible local and inter-area connectivity, guided by experimental observations of connectivity of the primate visual cortex. In our model, all information transfer is biologically motivated, inference and learning occur without phases, and network dynamics demonstrably approximate those of error backpropagation. We show the capabilities of our model on a wide range of benchmarks, and compare to other models, such as dendritic hierarchical predictive coding. In particular, our model addresses shortcomings of other theories in terms of scalability to many cortical areas. Finally, we make concrete predictions, which differentiate it from other theories, and which can be tested experimentally.Coherent cross-modal generation of synthetic biomedical data to advance multimodal precision medicine
by Raffaele Marchesi, Nicolò Lazzaro, Walter Endrizzi, Gianluca Leonardi, Matteo Pozzi, Flavio Ragni, Stefano Bovo, Monica Moroni, Venet Osmani, Giuseppe Jurman
Integration of multimodal, multi-omics data is critical for advancing precision medicine, yet its application is frequently limited by incomplete datasets where one or more modalities are missing. To address this challenge, we developed a generative framework capable of synthesizing any missing modality from an arbitrary subset of available modalities. We introduce Coherent Denoising, a novel ensemble-based generative diffusion method that aggregates predictions from multiple specialized, single-condition models and enforces consensus during the sampling process. We compare this approach against a multi-condition, generative model that uses a flexible masking strategy to handle arbitrary subsets of inputs. The results show that our architectures successfully generate high-fidelity data that preserve the complex biological signals required for downstream tasks. We demonstrate that the generated synthetic data can be used to maintain the performance of predictive models on incomplete patient profiles and can leverage counterfactual analysis to guide the prioritization of diagnostic tests. We validated the framework’s efficacy on a large-scale multimodal, multi-omics cohort from The Cancer Genome Atlas (TCGA) of over 10,000 samples spanning across 20 tumor types, using data modalities such as copy-number alterations (CNA), transcriptomics (RNA-Seq), proteomics (RPPA), and histopathology (WSI). This work establishes a robust and flexible generative framework to address sparsity in multimodal datasets, providing a key step toward improving precision oncology.Developmental and aging changes in brain network switching dynamics revealed by EEG phase synchronization
by Dionysios Perdikis, Rita Sleimen-Malkoun, Viktor Müller, Viktor Jirsa
Adaptive behavior depends on the brain’s capacity to vary its activity across multiple spatial and temporal scales. Yet, how distinct facets of this variability evolve from childhood to older adulthood remains poorly understood, limiting mechanistic models of neurocognitive aging. Here, we characterize lifespan neural variability using an integrated empirical-computational approach. We analyzed high-density EEG cohort data spanning 111 healthy individuals aged 9–75 years, recorded at rest and during passive and attended auditory oddball stimulation task. We extracted scale-dependent measures of EEG fluctuations amplitude and entropy, together with millisecond-resolved phase-synchrony networks in the 2–20 Hz range. Multi-condition partial least squares decomposition analysis revealed two independent lifespan trajectories. First, slow-frequency power, variance and complexity at longer timescales declined monotonically with age, indicating a progressive dampening of low-frequency fluctuations and large-scale coherence. Second, the temporal organization of phase-synchrony reconfigurations followed an inverted U-trend: young adults exhibited the slowest yet most diverse switching—characterized by low mean but high variance and low kurtosis of jump lengths at 2–6 Hz and the opposite pattern at 8–20 Hz—whereas children and older adults showed faster, more stereotyped dynamics. To mechanistically account for these patterns, we fitted a ten-node phase-oscillator model constrained by the human structural connectome. Only an intermediate, metastable coupling regime reproduced qualitatively the empirical finding of maximally heterogeneous synchrony dynamics observed in young adults, whereas deviations toward weaker or stronger coupling mimicked the children’s and older adults’ profiles. Our results demonstrate that development and aging entail changes in the switching dynamics of EEG phase synchronization, by differentially sculpting stationary and transient aspects of neural variability. This establishes time-resolved phase-synchrony metrics as sensitive, mechanistically grounded markers of neurocognitive status across the lifespan.Forecastability of infectious disease time series: are some seasons and pathogens intrinsically more difficult to forecast?
by Lauren A. White, Tomás M. León
For infectious disease forecasting challenges, individual model performance typically varies across space and time. This phenomenon raises the question: are there properties of the target time series that contribute to a particular season, location, or disease being more difficult to forecast? Here we characterize a time series’ future predictability using a forecastability metric that calculates the spectral entropy of the time series. Forecastability of syndromic influenza hospital admissions for the state of California varied widely across seasons and was positively correlated with peak burden. Next, using archived U.S. state and national forecasts targeting laboratory-confirmed COVID-19 and influenza hospital admissions, we investigated the relationship between forecastability and: (i) population size of the forecasting target, and (ii) forecast performance as measured by mean absolute error, weighted interval score (WIS), and scaled relative WIS. Forecastability increased with increasing population size of the forecasting target, and forecasting performance generally improved with higher forecastability when mitigating the effects of population size across scales. These preliminary results support the idea that some targets and respiratory virus seasons may be inherently more difficult to forecast and could help explain inter-seasonal variation in model performance.How muscle ageing affects rapid goal-directed movement: mechanistic insights from a simple model
by Delyle T. Polet, Christopher T. Richards
As humans and other animals age, passive and active muscle properties change markedly, with reduced peak tension, peak strain rate, activation and deactivation rate, and increased parallel stiffness. It is thought that these alterations modify locomotor performance, but establishing causal links is difficult when many parameters vary at once. We developed a simplified model of an elbow joint with two antagonistic Hill-type muscles, and varied the associated muscle parameters combinatorially over a large range. For a given parameter combination, we found optimal joint movements that minimized cumulative squared error to a target while starting and ending at rest. Emergent behaviour from the optimisations compared well to ballistic point-to-point arm movements in humans. Age-associated reductions of maximum isometric force, maximum strain rate and activation rate all had detrimental effects on performance, independent of other parameters. In contrast, deactivation time and passive parallel stiffness had no effect on performance on their own, but pronounced interactive effects with each other. Increasing stiffness reduced joint movement time at fast deactivation rates, but increased movement time at slow deactivation rates. This occurs because antagonist muscles resist the passive tension at rest, but are stretched eccentrically by the agonist, amplifying their active resistive force. Fast-deactivating muscles can avoid this resistive effect, allowing the passive stiffness to amplify accelerating force and enhance performance. In all cases, coactivation emerged as optimal during and after the braking period, and during the acceleration phase when stiffness increased. As deactivation time increased, so too did coactivation levels– but coactivation was not generally associated with a reduction in performance. Our simulations offer evidence that age-related changes in muscle strength, activation time and maximum contraction velocity can reduce ballistic performance in a goal-directed task, but the effects of increased muscle stiffness and deactivation time depend on their relative values.