Biology subjects

Richardson, M.

Publications and source records attributed to Richardson, M..

At least 19 recordsLinked to original sources

bioRxiv2026

Topology-Based Query Framework for Longitudinal Omics Trajectories

1 Abstract 1.1 Background Many longitudinal omics studies contain only a small number of repeated measurements collected before, during, or after an intervention. Existing approaches, including mixed-effects models and generalized additive models, estimate temporal effects but do not generally provide a discrete representation of trajectory topology that can be queried directly across experimental groups. 1.2 Methods We developed LongOmicsTraj, an open-source R package for topology-based representation and querying of short longitudinal omics trajectories. The framework encodes the direction of change between adjacent visits as up, down, or flat, with the ordered sequence defining an Ordinal Trajectory State (OTS). LongOmicsTraj operates downstream of trajectory estimation and can therefore be applied to empirical summaries or model-derived visit-level estimates, including those from linear mixed-effects models, generalized additive models, and polynomial regression, following a maSigPro-style time-course formulation [1]. OTS labels provide a common representation for topology-based querying, cross-group comparison, and evaluation of higherlevel representations such as trajectory clusters. We evaluated the framework using controlled simulations and bronchial biopsy transcriptomic data from the GLUCOLD corticosteroid intervention study (GEO accession GSE36221), measured at baseline, 6 months, and 30 months. The biological analysis compared continued inhaled corticosteroid (ICS) treatment, ICS withdrawal after 6 months, and placebo. 1.3 Results In simulations, LongOmicsTraj recovered predefined stable, monotonic, transient, rebound, and oscillatory trajectories with high accuracy when longitudinal signal was sufficiently clear, with performance declining under high-noise conditions and depending partly on the upstream estimator. In GLUCOLD, comparator-aware topology queries reduced 20,358 measured transcripts to 168 genes showing a corticosteroid response that was maintained during continued treatment, reversed following withdrawal, and was not reproduced under placebo. The selected genes included established corticosteroid-response genes and were enriched for immune-cell migration, chemotaxis, cell adhesion, and extracellular-matrix organisation. Topology-aware evaluation of FlexMix trajectory clusters additionally revealed substantial within-cluster temporal heterogeneity, with topology purities of approximately 46% to 60%. 1.4 Conclusions LongOmicsTraj provides a compact, directly queryable representation of temporal direction and order in short longitudinal omics studies. It complements existing longitudinal estimation and clustering methods by making trajectory structure explicit, enabling structured cross-group queries and quantification of temporal heterogeneity within trajectory clusters.

bioinformatics↗

bioRxiv2026

Automated Segmentation of Prostatic Gold Fiducial Markers for MR-Only Radiotherapy Planning Using Multi-Modal Consensus Deep Learning

PurposeTo develop and evaluate a multi-model consensus deep learning approach for automated gold fiducial marker (FM) segmentation in T1-weighted prostate MRI. Materials and MethodsIn this retrospective study, T1-weighted MRI and CT-derived reference standard segmentations were collected from 127 prostate cancer patients (all male; mean age, 70 years {+/-} 7 [standard deviation]; age range, 50-88 years; collected between October 2020 and January 2026) who each had three implanted gold FMs. A 3D U-Net was trained on 93 subjects using four random seeds to produce an ensemble. At inference, marker-class probability maps were averaged across models and the top three connected components selected. Performance was evaluated on 34 temporally held-out subjects (9 tuning, 25 test) using marker-level sensitivity and precision with exact (Clopper-Pearson) 95% confidence intervals (CIs). A model count ablation study was performed. The pipeline was deployed for on-scanner processing on Siemens MRI systems via the OpenRecon framework and as a browser-based application using WebAssembly, executing entirely client-side. ResultsThe four-model consensus achieved 96% (70 of 73) sensitivity and 95% (70 of 74) precision on 25 test subjects, with 29 of 34 (85%) subjects achieving perfect marker detection. Single models had a mean sensitivity of 84% (SD, 9%), improving to 96% with four-model consensus (SD, <1%). ConclusionMulti-model consensus deep learning substantially improved FM segmentation reliability over individual models, achieving high sensitivity and precision using only routinely acquired T1-weighted MRI.

bioinformatics↗

bioRxiv2026

Building an Interoperable Rare Disease Multi-omic Resource: The GREGoR Data Model and Dataset

Rare disease research and diagnosis rely on the integration of genomic and phenotypic data generated across diverse clinical sites; however, the absence of widely adopted standards for representing genomic data and associated metadata has limited data interoperability, reuse, and cross-study analysis. The Genomics Research to Elucidate the Genetics of Rare Diseases (GREGoR) Consortium was established to investigate challenging rare disease cases and evaluate emerging multi-omic technologies for clinical translation. To support coordinated data integration across distributed research sites, we developed a common Consortium Data Model in partnership with domain experts to standardize the capture of participant-, family-, phenotype- and assay-level metadata, with a particular emphasis on using a modular architecture to support linking of multiple data versions from multiple omic technologies to a single individual and attribution of a genetic finding to the specific technology used for its initial discovery. Adoption of the GREGoR Data Model has enabled continued generation and public release of a harmonized, analysis-ready Consortium Dataset. The most recent release includes phenotypic, family and multi-omic data from 12,292 participants in 5,029 families. Other rare disease data sharing efforts are beginning to adopt this data model which will facilitate cross consortium analyses and empower rare disease research. This work demonstrates that a collaborative, flexible, and scalable data model can enable large-scale rare disease research, facilitate cross-center data harmonization, and enable data interoperability.

genomics↗

bioRxiv2026

Long-read MitoScope reveals tissue-resolved somatic mitochondrial variation and landscape of nuclear-embedded mitochondrial sequences

The mitochondrial genome (mtDNA), rich in repeats and prone to nuclear mitochondrial DNA segments (NUMTs), drives somatic mosaicism implicated in cancer, metabolic syndromes, and neurodegeneration, yet short-read sequencing yields incomplete catalogs, mapping artifacts, and false heteroplasmies. Here, we introduce MitoScope, a scalable long-read workflow to assemble mtDNA, perform high-fidelity variant calling, resolve heteroplasmy, and characterize NUMTs in benchmarking tissues from the Somatic Mosaicism Across Human Tissues (SMaHT) Network. MitoScope shows high sensitivity and precision, determines copy number, and uncovers low-frequency variants. We define an age- and tissue-dependent landscape of mtDNA mosaicism, including low-frequency pathogenic heteroplasmies, a bimodal heteroplasmy spectrum shaped by purifying selection, and age-accumulating deletions enriched for microhomology. Parallel profiling of NUMTs identifies high-confidence events with >2-fold more NUMTs than short-read surveys--with evidence of nonrandom trinucleotide contexts at breakpoints. These findings expose pervasive, tissue-resolved somatic mtDNA and NUMT instability with direct relevance for variant interpretation, aging, and human disease.

genomics↗

bioRxiv2026

A human subcortical connectome at 400 μm resolution

The fiber pathways of the human subcortex are of particular importance in clinical neuroscience, as they are targeted by neuromodulation therapies, such as deep brain stimulation (DBS) and lesion approaches, in various motor and psychiatric disorders. However, the complexity, size, and position of these pathways make them challenging to image non-invasively with diffusion MRI (dMRI). As a result, recent efforts in atlasing these pathways to guide neuromodulation have resorted to using synthetic data. Here we present the first extensive reconstruction of fiber pathways of the human subcortex with ultra-high-resolution dMRI. We leverage a unique, ex vivo dMRI dataset acquired by the BRAIN CONNECTS center for Large-scale Imaging of Neural Circuits (LINC) on the first-of-its-kind Connectome 2.0 scanner. Our contribution is two-fold. First, by showing the feasibility of reconstructing these pathways with non-invasive neuroimaging at the single subject level, we set the stage for future research into reconstructing them in vivo in individual patients. Second, we provide a high-definition atlas of basal-ganglia-thalamocortical circuits that is readily extensible via our publicly released data and annotations. As a first demonstration of its clinical validity, we align our atlas to "hotspots" from previous DBS studies and identify pathways associated with therapeutic or side effects.

neuroscience↗

bioRxiv2026

Face-selective responses correlate with deep networks that learn from environment feedback

AO_SCPLOWBSTRACTC_SCPLOWDeep neural networks have shown high accuracy in modeling neural responses in the visual system, but most models rely on supervised learning, which requires training on ground-truth labels that are typically unavailable in real-world settings. While unsupervised models can address this limitation, they miss another key aspect: visual representations are shaped by feedback from the environment. We introduce a reinforcement learning (RL) model of face perception that incorporates both input stimuli and feedback from the environment. Inspired by human interactions, we train the model to approach faces yielding positive interactions and avoid faces yielding negative interactions. Using intracortical electroencephalography (iEEG) data and Representational Dissimilarity Matrices (RDMs), we evaluate the models ability to account for neural responses. Our RL model performs at the same level as supervised and unsupervised models, capturing neural responses to complex visual stimuli. The findings suggest that RL models are a promising approach for understanding perception. Significance StatementUnderstanding how the brain encodes faces is central to vision science. Existing models rely on supervised learning, which requires ground-truth labels that are often unavailable in real-world settings, or on unsupervised learning, which ignores the role of environmental-feedback in shaping visual representations. We introduce a reinforcement learning (RL) model that learns through environmental feedback, simulating human interactions by associating approaching faces with positive interactions and avoiding faces with negative interactions. Using intracortical electroencephalography (iEEG) data from face-selective regions, we show that an RL model with a variational DenseNet encoder accounts for neural representations comparably to supervised and unsupervised models. Task and architecture jointly shaped representational geometry, highlighting the importance of both learning objective and encoder design. These findings suggest the potential of RL-based approaches to understand neural representations of naturalistic faces.

neuroscience↗

bioRxiv2026

Donor-specific assemblies enhance somatic structural variant detection in complex genomic regions

Structural variants (SVs) contribute substantially to genomic variation and disease, but detecting somatic SVs (sSVs) remains difficult due to reference bias, mosaicism, and enrichment in repetitive regions. Linear reference genomes, like GRCh38 and CHM13, do not fully capture individual genomic structure, which can obscure true somatic variation. Donor-specific assemblies (DSAs) generated from the same genome where sSVs are being assayed provide a personalized alternative, yet their performance for sSV detection has not been systematically assessed. As part of the Somatic Mosaicism across Human Tissues (SMaHT) Network, we benchmark a DSA for sSV discovery in the COLO829 melanoma cell line with a matched normal sample from the same individual. We compare sSV detection across GRCh38, CHM13, and the COLO829BL_DSA using three different sSV callers (Delly, Severus, and Sniffles2) and sequence data from multiple long-read platforms. The COLO829BL_DSA identifies 1.8-fold more manually validated sSVs than linear references, in regions both shared with GRCh38 and CHM13 and unique to the COLO829BL_DSA. Variants detected only with the COLO829BL_DSA are often found in satellite and other repeat-rich regions that are difficult to resolve using standard references. In addition, several COLO829BL_DSA-specific sSVs are located in genes, some of which are associated with cancer. Overall, these results underscore the utility of DSAs in improving sSV detection.

genomics↗

bioRxiv2026

mRNAfold: Co-optimization of Global Stability, Local Structure, and Codon Choice via Suboptimal Folding

mRNA medicines hold great promise, but designing sequences with high translation efficiency, robust in-solution stability, and manufacturability remains a major challenge due to the vast combinatorial space of synonymous coding sequences. Computational approaches such as mRNA folding algorithms have emerged as powerful tools by co-optimizing for in-solution stability and translation efficiency, yet current methods face important limitations. Here, we present "mRNAfold", an improved mRNA folding algorithm and software package that addresses these gaps by enabling efficient exploration of diverse near-optimal solutions, incorporating untranslated regions (UTRs), parallel execution, and supporting tunable control over local structural features across the mRNA. Thermodynamically optimized mRNAs from mRNAfold were more stable ({approx} 2-fold) in-solution than those generated by simple GC maximization for the same encoded protein. In addition, mRNAs designed to vary local structure near the start codon while maintaining consistent structure and codon optimality elsewhere showed a complex relationship between local structure near the start codon and protein production in cells. We observed no impact of structure in the start codon region for a set of mRNAs with high codon optimality, but it did impact protein production for a set of mRNAs with lower codon optimality. Together, these results underscore the potential of structure-aware, multi-objective design to improve mRNA medicines and offer a framework for exploring how sequence, structure, and expression are interrelated.

bioinformatics↗

bioRxiv2025

Large-scale culturing of the tree microbiome enables targeted disease suppression

The tree microbiome is essential for host health and pathogen suppression. Synthetic microbial communities (SynComs) are emerging as important tools to understand microbiome dynamics and engineer microbiomes to harness beneficial properties. However, while the rational design, assembly and application of SynComs requires representative microbiota isolate collections combined with functional information, microbial culture collections from tree species such as oak (Quercus) are critically lacking. Here, we generated an oak microbiota culture collection comprising >30,000 isolates from 150 oak trees across Britain, belonging to key bacterial and fungal taxa that represented 61% of the total bacterial sequences and 87% of total fungal sequences as determined by culture-independent sequencing. Over 22,000 isolates were screened for suppression of bacterial species associated with degradation of live stem tissue in trees affected by Acute Oak Decline (AOD), identifying 341 bacterial isolates that suppressed oak pathogens. In vitro screening of 40 randomly assembled SynComs demonstrated that oak microbiota SynComs can suppress oak pathogenic bacteria associated with AOD. Inoculation of a disease-suppressive SynCom into the stem of oak seedlings and logs prior to pathogen challenge reduced the quantities of the bacteria, Brenneria goodwinii and Gibbsiella quercinecans, by 56% and 87%, respectively, in seedlings, and 71% and 95% in logs. This work demonstrates that the tree microbiome can be engineered using disease suppressive SynComs.

microbiology↗

bioRxiv2025

A hierarchical cascade of sleep rhythms drive memory consolidation in humans and are disrupted in epilepsy.

The cross-regional interplay of slow oscillations, sleep spindles, and ripples during sleep is believed to support systems memory consolidation but is understudied in humans. Using a validated behavioral task and intracranial neural recordings from orbitofrontal cortex, thalamus, and hippocampus in 19 epilepsy patients, we examined the cross-regional interplay of sleep-oscillations and their role in memory consolidation. Orbitofrontal slow oscillations robustly modulate sleep rhythms both within and across regions. Most combinations of oscillation rates predict overnight memory consolidation, but hippocampal ripple rate and coupled hippocampal-orbitofrontal ripples were the strongest positive predictors of memory consolidation. In contrast, epileptic spikes coupled to sleep oscillations strongly predicted reduced memory consolidation, with the strongest negative effect observed when epileptic spikes were coupled to slow oscillations. These findings provide direct evidence of the hierarchical cascade of sleep oscillations in human memory processing and reveal how epileptic spikes disrupt this process in patients with epilepsy.

neuroscience↗

bioRxiv2025

Copper supplementation mitigates Parkinson-like wild-type SOD1 pathology and nigrostriatal degeneration in a novel mouse model

Wild-type superoxide dismutase 1 (disSOD1) protein misfolding and deposition is implicated in the death of substantia nigra (SN) dopamine neurons in Parkinson disease. Regionally reduced copper availability, and subsequent reduced copper binding to SOD1, is a key factor driving the development of this pathology, suggesting brain copper supplementation may constitute an effective means of preventing its formation. We evaluated the potential of the blood-brain-barrier-permeable copper delivery drug, CuATSM, to attenuate the misfolding and deposition of wild-type disSOD1, and associated neuron death, in a novel mouse model that expresses this pathology. Using proteomic and elemental mass spectrometry, together with biochemical and histological workflows, we demonstrated copper supplementation corrects altered post-translational modifications on soluble SOD1 and improves the enzymatic activity of the protein in the brains of these animals. These changes were associated with a significant reduction in disSOD1 pathology and preservation of dopamine neurons in the SN, which were highly correlated with tissue copper levels. These data position wild-type disSOD1 pathology as a novel drug target for Parkinson disease and suggest that brain copper supplementation may constitute an effective means of slowing SN dopamine neuron death in this disorder.

neuroscience↗

bioRxiv2025

Zero-phase-delay connectivity increases the reliability, concordance with structure, and prognostic ability of functional connectivity metrics

Neural populations synchronise their activity with either zero-phase-delay (activity in interacting regions occurs simultaneously) or a phase-delay (activity in one region follows the other). In electroencephalography and magnetoencephalography functional connectivity analyses, artefactual connectivity can also occur with zero-phase-delay. To minimise artefact, contemporary analyses typically exclude all zero-phase-delay interactions. However, the extent to which true interactions are resultingly lost - and the impact this has on the performance of functional connectivity metrics as biomarkers - remains unknown. Here, we show that most cortico-cortical functional connectivity occurs with zero- or near-zero phase-delay, even where such connectivity is unlikely to be artefactual. Including, rather than excluding, zero-phase-delay connectivity increases the reliability, convergence with neurobiology (structure-function concordance, homotopic interhemispheric connectivity, and age-related connectivity changes), and prognostic ability of functional connectivity metrics. We find that excluding zero-phase-delay connections penalises functional connectivity strength between the strongest structurally connected regions: stronger structural connections lead to functional connections with phase-delays closer to zero, mediated by a shorter signal propagation time. Our findings challenge generally accepted assumptions that zero-phase-exclusive methods are superior to zero-phase-inclusive methods.

neuroscience↗

bioRxiv2024

Personalized cue-reactive delta-theta oscillations guide deep brain stimulation for opioid use disorder

Abstract / SummarySubstance use disorder (SUD) is a significant public health concern, with over 30% of the affected population not responding to available treatments. Severe SUD is characterized by drug-cue reactivity that has been reported to predict treatment-failure. We leveraged this pathophysiological feature to optimize deep brain stimulation (DBS) of the nucleus accumbens region (NAc) in an adult with SUD. A personalized drug cue-reactivity task was administered while recording NAc region electrophysiology from a lead externalized for clinical purposes. We identified a drug cue-evoked signal in the ventral NAc associated with intensification of opioid-related cravings, which attenuated subsequent to stimulation delivered to the same area. DBS was then programmed to engage this focal region, which resulted in sustained suppression of drug-related cravings. This finding heralds the potential for personalized strategies to optimize DBS for SUD.

neuroscience↗

bioRxiv2024

Universality of representation in biological and artificial neural networks

Many artificial neural networks (ANNs) trained with ecologically plausible objectives on naturalistic data align with behavior and neural representations in biological systems. Here, we show that this alignment is a consequence of convergence onto the same representations by high-performing ANNs and by brains. We developed a method to identify stimuli that systematically vary the degree of inter-model representation agreement. Across language and vision, we then showed that stimuli from high-and low-agreement sets predictably modulated model-to-brain alignment. We also examined which stimulus features distinguish high-from low-agreement sentences and images. Our results establish representation universality as a core component in the model-to-brain alignment and provide a new approach for using ANNs to uncover the structure of biological representations and computations.

neuroscience↗

bioRxiv2024

MCSPACE: inferring microbiome spatiotemporal dynamics from high-throughput co-localization data

Recent advances in high-throughput approaches for estimating co-localization of microbes, such as SAMPL-seq, allow characterization of the biogeography of the gut microbiome longitudinally and at unprecedented scale. However, these high-dimensional data are complex and have unique noise properties. To address these challenges, we developed MCSPACE, a probabilistic AI method that infers from microbiome co-localization data spatially coherent assemblages of taxa, their dynamics over time, and their responses to perturbations. To evaluate MCSPACEs capabilities, we generated the largest longitudinal microbiome co-localization dataset to date, profiling spatial relationships of microbes in the guts of mice subjected to serial dietary perturbations over 76 days. Analyses of these data and an existing human longitudinal dataset demonstrated superior benchmarking performance of MCSPACE over existing methods, and moreover yielded insights into spatiotemporal structuring of the gut microbiome, including identifying temporally persistent and dynamic microbial assemblages in the human gut, and shifts in assemblages in the murine gut induced by specific dietary components. Our results highlight the utility of our method, which we make available to the community as an open-source software tool, for elucidating dynamics of microbiome biogeography and gaining insights into the role of spatial relationships in host-microbial ecosystem function.

microbiology↗

bioRxiv2024

Impaired maturation of wild-type superoxide dismutase 1 associated with neurodegeneration in Parkinson disease brain and a novel murine model

Atypical wild-type superoxide dismutase 1 (SOD1) protein misfolding and deposition occurs specifically within the degenerating substantia nigra pars compacta (SNc) in Parkinson disease. Mechanisms driving the formation of this pathology and relationship with SNc dopamine neuron health, are yet to be fully understood. We applied proteomic mass spectrometry and synchrotron-based biometal quantification to post-mortem brain tissues from the SNc of Parkinson disease patients and age-matched controls to uncover key factors underlying the formation of wild-type SOD1 pathology in this disorder. We then engineered brain copper deficiency and upregulated SOD1 protein levels in a novel mouse strain, termed the SOCK mouse, to verify their involvement in the development of Parkinson-like wild-type SOD1 pathology and their impact on dopamine neuron health. Soluble SOD1 protein in the degenerating Parkinson disease SNc exhibited altered post-translational modifications, which may underlie changes to the enzymatic activity and aggregation of the protein in this region. These include decreased copper binding, dysregulation of physiological glycosylation, and atypical oxidation and glycation of key SOD1 amino acid residues. We demonstrated that the biochemical profile introduced in SOCK mice promotes the same post-translational modifications and the development of Parkinson-like wild-type SOD1 pathology in the midbrain and cortex. This pathology accumulates progressively with age and is accompanied by nigrostriatal degeneration and dysfunction, which occur in the absence of -synuclein deposition. These mice do not exhibit weight loss nor spinal cord motor neuron degeneration, distinguishing them from transgenic mutant SOD1 mouse models. This study provides the first in vivo evidence that mismetallation and altered post-translational modifications precipitates wild-type SOD1 misfolding, dysfunction and deposition in the Parkinson disease brain, which may contribute to SNc dopamine neuron degeneration. Our data position this pathology as a novel drug target for this disorder, with a particular focus on therapies capable of correcting alterations to SOD1 post-translational modifications.

neuroscience↗

bioRxiv2024

SAMPL-seq reveals micron-scale spatial hubs in the human gut microbiome

The local arrangement of microbes can profoundly impact community assembly, function, and stability. To date, little is known about the spatial organization of the human gut microbiome. Here, we describe a high-throughput and streamlined method, dubbed SAMPL-seq, that samples microbial composition of micron-scale sub-communities with split-and-pool barcoding to capture spatial colocalization in a complex consortium. SAMPL-seq analysis of the gut microbiome of healthy humans identified bacterial taxa pairs that consistently co-occurred both over time and across multiple individuals. These colocalized microbes organize into spatially distinct groups or "spatial hubs" dominated by Bacteroideceae, Ruminococceae, and Lachnospiraceae families. From a dietary perturbation using inulin, we observed reversible spatial rearrangement of the gut microbiome, where specific taxa form new local partnerships. Spatial metagenomics using SAMPL-seq can unlock new insights to improve the study of microbial communities. One Sentence SummaryHigh throughput micron-scale subcommunity sampling and sequencing identifies distinct spatial associations of gut bacteria within and across individuals.

systems biology↗

bioRxiv2024

Spontaneous HFO Sequences Reveal Propagation Pathways for Precise Delineation of Epileptogenic Networks

AbstractEpilepsy, a neurological disorder affecting millions worldwide, poses great challenges in precisely delineating the epileptogenic zone - the brain region generating seizures - for effective treatment. High-frequency oscillations (HFOs) are emerging as promising biomarkers; however, the clinical utility is hindered by the difficulties in distinguishing pathological HFOs from non- epileptiform activities at single electrode and single patient resolution and understanding their dynamic role in epileptic networks. Here, we introduce an HFO-sequencing approach to analyze spontaneous HFOs traversing cortical regions in 40 drug-resistant epilepsy patients. This data- driven method automatically detected over 8.9 million HFOs, pinpointing pathological HFO- networks, and unveiled intricate millisecond-scale spatiotemporal dynamics, stability, and functional connectivity of HFOs in prolonged intracranial EEG recordings. These HFO sequences demonstrated a significant improvement in localization of epileptic tissue, with an 818.47% increase in concordance with seizure-onset zone (mean error: 2.92 mm), compared to conventional benchmarks. They also accurately predicted seizure outcomes for 90% AUC based on pre-surgical information using generalized linear models. Importantly, this mapping remained reliable even with short recordings (mean standard deviation: 3.23 mm for 30-minute segments). Furthermore, HFO sequences exhibited distinct yet highly repetitive spatiotemporal patterns, characterized by pronounced synchrony and predominant inward information flow from periphery towards areas involved in propagation, suggesting a crucial role for excitation-inhibition balance in HFO initiation and progression. Together, these findings shed light on the intricate organization of epileptic network and highlight the potential of HFO-sequencing as a translational tool for improved diagnosis, surgical targeting, and ultimately, better outcomes for vulnerable patients with drug-resistant epilepsy. One Sentence SummaryPathological fast brain oscillations travel like traffic along varied routes, outlining recurrently visited neural sites emerging as critical hotspots in epilepsy network.

neuroscience↗