The maintenance of diversity in complex ecological communities despite unpredictable dynamics and competitive exclusion is thought to require continual influx of new species or competitive advantages that accrue as species become rare. We examine isolated planktonic microbial communities under permanent ice cover in Antarctic lakes, recording prokaryotic abundance across 9 communities, 11 years, 30~m of depth, and thousands of species in the McMurdo LTER. We quantify rare species advantage by modeling community dynamics under frequency-dependent selection. We find persistent diversity and pervasive negative frequency dependence with limited immigration and turnover. While ecology and evolutionary sciences have long debated whether diversity is maintained selectively, we measure selection over a $10^4$-fold range of abundance in naturally coevolving communities and implicate rare species advantage.

Mooney images are high-contrast, two-tone visual stimuli, created by thresholding photographic images. They allow researchers to separate image content from image understanding, making them valuable for studying visual perception. An ideal Mooney image for this purpose achieves a specific balance: it initially appears unrecognizable but becomes fully interpretable to the observer after seeing the original template. Researchers traditionally created these stimuli manually using subjective criteria, which is labor-intensive and can introduce inconsistencies across studies. Automated generation techniques now offer an alternative to this manual approach. Here, we present MooneyMaker, an open-source Python package that automates the generation of ambiguous Mooney images using several complementary approaches. Users can choose between various generation techniques that range from approaches based on image statistics to deep learning models. These models strategically alter edge information to increase initial ambiguity. The package lets users create two-tone images with multiple methods and directly compare the results visually. In an experiment, we validate MooneyMaker by generating Mooney images using different techniques and assess their recognizability for human observers before and after disambiguating them by presenting the template images. Our results reveal that techniques with lower initial recognizability are associated with higher post-template recognition (i.e. a larger disambiguation effect). To help vision scientists build effective databases of Mooney stimuli, we provide practical guidelines for technique selection. By standardizing the generation process, MooneyMaker supports more consistent and reproducible visual perception research.

Component systems - ensembles of realizations built from a shared repertoire of modular parts - are ubiquitous in biological, ecological, technological, and socio-cultural domains. From genomes to texts, cities, and software, these systems exhibit statistical regularities that often meet the "bona fide" requirements of laws in the physical sciences. Here, we argue that the generality and simplicity of those laws are often due to basic combinatorial or sampling constraints, raising the question of whether such patterns are actually revealing system-specific mechanisms and how we might move beyond them. To this end, we first present a unifying mathematical framework, which allows us to compare modular systems in different fields and highlights the common "null" trends as well as the system-specific uniqueness, which, arguably, are signatures of the underlying generative dynamics. Next, we can exploit the framework with statistical mechanics and modern machine-learning tools for a twofold objective. (i) Explaining why the general regularities emerge, highlighting the constraints between them and the general principles at their origins, and (ii) "subtracting" them from data, which will isolate the informative features for inferring hidden system-specific generative processes, mechanistic and causal aspects.

Accurate, low-latency estimates of the instantaneous phase of oscillations are essential for closed-loop sensing and actuation, including (but not limited to) phase-locked neurostimulation and other real-time applications. The endpoint-corrected Hilbert transform (ecHT) reduces boundary artefacts of the Hilbert transform by applying a causal narrow-band filter to the analytic spectrum. This improves the phase estimate at the most recent sample. Despite its widespread empirical use, the systematic endpoint distortions of ecHT have lacked a principled, closed-form analysis. In this study, we derive the ecHT endpoint operator analytically and demonstrate that its output can be decomposed into a desired positive-frequency term (a deterministic complex gain that induces a calibratable amplitude/phase bias) and a residual leakage term setting an irreducible variance floor. This yields (i) an explicit characterisation and bounds for endpoint phase/amplitude error, (ii) a mean-squared-error-optimal scalar calibration (c-ecHT), and (iii) practical design rules relating window length, bandwidth/order, and centre-frequency mismatch to residual bias via an endpoint group delay. The resulting calibrated ecHT achieves near-zero mean phase error and remains computationally compatible with real-time pipelines. Code and analyses are provided at https://github.com/eosmers/cecHT.

Bacteria regulate their motility through a variety of mechanisms, including quorum sensing (QS) and other density-dependent responses mediated by diffusible signals. While nonlinear density-dependent motility is well known in active-matter theory to generate nonequilibrium spatial patterns, its consequences for the coexistence of growing, interacting species remain less explored. Here we develop a minimal spatially structured model for two strongly competing species in which local demographic interactions are coupled to an escape response: each species increases its motility nonlinearly (sigmoidal) with the local abundance of its competitor. We show that this sigmoidal motility regulation promotes optimal spatial self-organization and can sustain long term coexistence via segregation, even in parameter regimes that yield competitive exclusion in well-mixed Lotka-Volterra dynamics. On two-dimensional lattices, the interplay between demographic competition and density-dependent motility generates a range of emergent patterns, including regimes in which the weaker competitor counterintuitively has higher total abundance. Overall, our results identify nonlinear, competitor-induced motility as a fundamental mechanism capable of sustaining coexistence in competing microbial populations.

Aortic valve opening (AO) events are crucial for detecting frequency and rhythm disorders, especially in real-world settings where seismocardiography (SCG) signals collected via consumer smartphones are subject to noise, motion artifacts, and variability caused by device heterogeneity. In this work, we present a robust deep-learning framework for SCG segmentation and rhythm analysis using accelerometer recordings obtained with consumer smartphones. We develop an enhanced U-Net v3 architecture that integrates multi-scale convolutions, residual connections, and attention gates, enabling reliable segmentation of noisy SCG signals. A dedicated post-processing pipeline converts probability masks into precise AO timestamps, whereas a novel adaptive 3D-to-1D projection method ensures robustness to arbitrary smartphone orientation. Experimental results demonstrate that the proposed method achieves consistently high accuracy and robustness across various device types and unsupervised data-collection conditions. Our approach enables practical, low-cost, and automated cardiac-rhythm monitoring using everyday mobile devices, paving the way for scalable, field-deployable cardiovascular assessment and future multimodal diagnostic systems.

Decoding visual semantic representations from human brain activity is a significant challenge. While recent zero-shot decoding approaches have improved performance by leveraging aligned image-text datasets, they overlook a fundamental aspect of human cognition: semantic understanding is inherently anchored in the auditory modality of speech, not text. To address this, our study introduces the first comparative framework for evaluating auditory versus textual semantic modalities in zero-shot visual neural decoding. We propose a novel brain-visual-auditory multimodal alignment model that directly utilizes auditory representations to encapsulate semantics, serving as a substitute for traditional textual descriptors. Our experimental results demonstrate that the auditory modality not only surpasses the textual modality in decoding accuracy but also achieves higher computational efficiency. These findings indicate that auditory semantic representations are more closely aligned with neural activity patterns during visual processing. This work reveals the critical and previously underestimated role of auditory semantics in decoding visual cognition and provides new insights for developing brain-computer interfaces that are more congruent with natural human cognitive mechanisms.

During infectious disease epidemics, pathogen transmission occurs in host populations made up of interacting subpopulations. Using stochastic simulation and analytical approximations, we examine how outbreak sizes in networked populations depend on network architecture, subpopulation sizes and the strength of coupling between subpopulations. We find, as expected, that mean outbreak sizes are frequently lower in networked populations than in homogeneous populations with the same basic reproduction number. However, after an outbreak ends, a networked population is often vulnerable to further outbreaks, and the ending of an outbreak may not imply herd immunity in any sense. Another key finding is that a relatively small amount of randomly distributed prior immunity can be more protective in a networked population than a homogeneous population, a phenomenon which can be reproduced analytically in certain cases. We also find that in networked populations, randomly distributed prior immunity is often more protective than infection-acquired immunity; but this conclusion can be reversed in populations with highly variable susceptibility. All of these conclusions have implications for designing outbreak control strategies that aim to reduce pathogen transmission during epidemics.

Hepatitis A severity increases sharply with age, while Korea is experiencing a cohort shift in which low seroprevalence adult cohorts are aging into older, higher fatality age groups. This demographic and immunological transition creates an urgent policy question regarding how adult vaccination should be prioritized under resource constraints. We evaluated three adult vaccination scenarios targeting low seroprevalence age groups (S1) 20 to 39 years, (S2) 40 to 59 years, and (S3) 20 to 59 years. Using an age structured dynamic transmission model calibrated to Korean data, we derived dynamically feasible vaccination allocation trajectories under realistic capacity constraints using an optimal control framework and linked these trajectories to long term transmission model simulations. We conducted DALY based cost effectiveness analyses over a lifetime horizon from both healthcare system and societal perspectives, and characterized uncertainty using probabilistic sensitivity analysis (PSA) and cost effectiveness acceptability curves (CEACs). Robustness was examined using one way sensitivity analyses. In the base case, S2 consistently yields the most favorable and robust cost effectiveness profile under both perspectives, with the lowest ICER. S3 achieved the largest reduction in DALYs but requires substantially higher incremental costs, resulting in a higher ICER than S2. S1 produces the smallest DALY reduction and is the least efficient strategy. PSA and CEACs confirm that S2 remains the preferred option across most willingness to pay ranges. S2 offers the most balanced and robustly cost effective strategy in Korea, capturing substantial mortality reduction while limiting additional program costs. S3 may be justified when higher budgets or willingness to pay thresholds are acceptable, but S2 provides the clearest value for money under epidemiological and economic conditions.

Identifying protein targets for small molecules, or reverse screening, is essential for understanding drug action, guiding compound repurposing, predicting off-target effects, and elucidating the molecular mechanisms of bioactive compounds. Despite its critical role, reverse screening remains challenging because accurately capturing interactions between a small molecule and structurally diverse proteins is inherently complex, and conventional step-wise workflows often propagate errors across decoupled steps such as target structure modeling, pocket identification, docking, and scoring. Here, we present an end-to-end reverse screening strategy leveraging HelixFold3, a high-accuracy biomolecular structure prediction model akin to AlphaFold3, which simultaneously models the folding of proteins from a protein library and the docking of small-molecule ligands within a unified framework. We validate this approach on a diverse and representative set of approximately one hundred small molecules. Compared with conventional reverse docking, our method improves screening accuracy and demonstrates enhanced structural fidelity, binding-site precision, and target prioritization. By systematically linking small molecules to their protein targets, this framework establishes a scalable and straightforward platform for dissecting molecular mechanisms, exploring off-target interactions, and supporting rational drug discovery.

Designing fluorescent small molecules with tailored optical and physicochemical properties requires navigating vast, underexplored chemical space while satisfying multiple objectives and constraints. Conventional generate-score-screen approaches become impractical under such realistic design specifications, owing to their low search efficiency, unreliable generalizability of machine-learning prediction, and the prohibitive cost of quantum chemical calculation. Here we present LUMOS, a data-and-physics driven framework for inverse design of fluorescent molecules. LUMOS couples generator and predictor within a shared latent representation, enabling direct specification-to-molecule design and efficient exploration. Moreover, LUMOS combines neural networks with a fast time-dependent density functional theory (TD-DFT) calculation workflow to build a suite of complementary predictors spanning different trade-offs in speed, accuracy, and generalizability, enabling reliable property prediction across diverse scenarios. Finally, LUMOS employs a property-guided diffusion model integrated with multi-objective evolutionary algorithms, enabling de novo design and molecular optimization under multiple objectives and constraints. Across comprehensive benchmarks, LUMOS consistently outperforms baseline models in terms of accuracy, generalizability and physical plausibility for fluorescence property prediction, and demonstrates superior performance in multi-objective scaffold- and fragment-level molecular optimization. Further validation using TD-DFT and molecular dynamics (MD) simulations demonstrates that LUMOS can generate valid fluorophores that meet various target specifications. Overall, these results establish LUMOS as a data-physics dual-driven framework for general fluorophore inverse design.

This research explores how total mortality figures relate to age-standardized death rates within the United States, using the complete historical record of national mortality statistics. Through a detailed investigation of both all-cause and cause-specific mortality trends, the study evaluates the impact of demographic standardization on interpreting mortality data across different time periods and geographic regions. Results indicate a robust and persistent association between crude death totals and age-adjusted rates. However, the findings also demonstrate that without adjusting for age, comparisons over time or across locations may misrepresent underlying epidemiological shifts, largely due to evolving population age structures. The study underscores the critical role of age adjustment as a methodological tool for generating accurate, interpretable, and comparable measures of public health outcomes.

The grandmother hypothesis is the most influential account of the evolution of menopause in humans, but other theories warrant investigation. Here I use simulations to investigate two theories that ground the evolution of menopause in biparental care. Kaplan et al. (2010) proposed a "two-sex" learning and skill-based account, termed the Embodied Capital Model (ECM), in which the high energetic burden of caring for multiple, slow-developing offspring was met with biparental investment. Menopause evolved because the physiological costs of pregnancy and childbirth increased with age yet productivity also increased with age, and the benefits of transferring resources to adult children and their offspring eventually outweigh the benefits of continued reproduction. Kuhle (2007) proposed the "father absent" hypothesis in which the higher mortality rate of husbands would often have left wives without the resources to raise young children, selecting for early reproductive cessation in monogamous couples. Simulations of hunter-gatherer energy consumption and production across the lifespan, taking account of age- and sex-specific survivorship, interbirth intervals, and varying rates of strength and foraging skill acquisition typical of contemporary foragers, reveal a pronounced midlife energy deficit that could be averted by ceasing reproduction midlife and receiving energy transfers from both younger couples (e.g., brideservice) and from older parents (the grandmother hypothesis). Menopause emerges as an integral and strictly necessary component of the unique human pattern of relatively short interbirth intervals and a long period of juvenile dependency, supporting and extending the ECM.

Existing methods in nanoscale connectomics are at present too slow to map entire mammalian brains. As an emerging approach, expansion microscopy (ExM) has enormous promise, yet it still suffers from throughput limitations. Mapping the human brain and even mapping nonhuman primate brains therefore remain distant goals. While ExM increases effective resolution linearly, it enlarges tissue volume cubically, which dramatically increases imaging time. As a rapid tomographic technique, X-ray microscopy has potential for drastically speeding up large-volume connectomics. But to the best of my knowledge, no group has so far imaged cellular features within expanded tissue using X-ray microscopy. I herein present an early-stage report featuring the first demonstration of X-ray microscopy reconstruction of cell bodies within expanded tissue. This was achieved by combining a modified enzymatic Unclearing technique with a metallic gold stain and imaging using a laboratory X-ray microscope. I emphasize that a great deal of work remains to develop "expansion X-ray microscopy" (ExXRM) to the point where it can be useful for connectomics since the current iteration of ExXRM only resolves cell bodies and not neurites due to extensive off-target staining. Additionally, the current method must be modified to accommodate for the challenges of synchrotron X-ray microscopy, a vastly speedier approach than laboratory X-ray microscopy. Nonetheless, achieving X-ray contrast in expanded tissues represents a significant first step towards realizing ExXRM as a connectomics imaging modality.

Understanding how the brain represents the multifaceted properties of words in context is essential for explaining the neural architecture of human language. Here, we combine large-scale psycholinguistic modeling with naturalistic fMRI to uncover the latent structure of word properties and their neural representations during narrative comprehension. By analyzing 106 psycholinguistic variables across 13,850 English words, we identified eight interpretable latent dimensions spanning lexical usage, word form, phonology orthography mapping, sublexical regularity, and semantic organization. These factors robustly predicted behavioral performance across lexical decision, naming, recognition, and semantic judgment tasks, demonstrating their cognitive relevance. Parcel-based and multivariate fMRI analyses of narrative listening revealed that these latent dimensions are encoded in overlapping yet functionally differentiated cortical systems. Multidimensional scaling and hierarchical clustering analyses further identified four interacting subsystems supporting sensorimotor grounding, controlled semantic retrieval, resolution of lexical competition, and contextual episodic integration. Together, these findings provide a unified neurocognitive framework linking fundamental lexical psycholinguistic dimensions to distributed cortical systems engaged during naturalistic language comprehension.

Scientific theories of consciousness should be falsifiable and non-trivial. Recent research has given us formal tools to analyze these requirements of falsifiability and non-triviality for theories of consciousness. Surprisingly, many contemporary theories of consciousness fail to pass this bar, including theories based on causal structure but also (as I demonstrate) theories based on function. Herein, I show these requirements of falsifiability and non-triviality especially constrain the potential consciousness of contemporary Large Language Models (LLMs) because of their proximity to systems that are equivalent to LLMs in terms of input/output function; yet, for these functionally equivalent systems, there cannot be any falsifiable and non-trivial theory of consciousness that judges them conscious. This forms the basis of a disproof of contemporary LLM consciousness. I then show a positive result, which is that theories of consciousness based on (or requiring) continual learning do satisfy the stringent formal constraints for a theory of consciousness in humans. Intriguingly, this work supports a hypothesis: If continual learning is linked to consciousness in humans, the current limitations of LLMs (which do not continually learn) are intimately tied to their lack of consciousness.

The paper proposes a modular-based approach to constraint handling in process optimization and control. This is partly motivated by the recent interest in learning-based methods, e.g., within bioproduction, for which constraint handling under uncertainty is a challenge. The proposed constraint handler, called predictive filter, is combined with an adaptive constraint margin and a constraint violation cost monitor to minimize the cost of violating soft constraints due to model uncertainty and disturbances. The module can be combined with any controller and is based on minimally modifying the controller output, in a least squares sense, such that constraints are satisfied within the considered horizon. The proposed method is computationally efficient and suitable for real-time applications. The effectiveness of the method is illustrated through a realistic case study of glycosylation constraint satisfaction in continuous monoclonal antibody biosimilar production using Chinese hamster ovary cells, employing a metabolic network model consisting of 23 extracellular metabolites and 126 reactions. In the case study, the average constraint-violation cost is reduced by more than 60% compared to the case without the proposed constraint-handling method.

Vision-Language Models (VLMs) have shown significant potential in surgical scene analysis, yet existing models are limited by frame-level datasets and lack high-quality video data with procedural surgical knowledge. To address these challenges, we make the following contributions: (i) SurgPub-Video, a comprehensive dataset of over 3,000 surgical videos and 25 million annotated frames across 11 specialties, sourced from peer-reviewed clinical journals, (ii) SurgLLaVA-Video, a specialized VLM for surgical video understanding, built upon the TinyLLaVA-Video architecture that supports both video-level and frame-level inputs, and (iii) a video-level surgical Visual Question Answering (VQA) benchmark, covering diverse 11 surgical specialities, such as vascular, cardiology, and thoracic. Extensive experiments, conducted on the proposed benchmark and three additional surgical downstream tasks (action recognition, skill assessment, and triplet recognition), show that SurgLLaVA-Video significantly outperforms both general-purpose and surgical-specific VLMs with only three billion parameters. The dataset, model, and benchmark will be released to enable further advancements in surgical video understanding.

Understanding the dynamics of excitation patterns in neural fields is an important topic in neuroscience. Neural field equations are mathematical models that describe the excitation dynamics of interacting neurons to perform the theoretical analysis. Although many analyses of neural field equations focus on the effect of neuronal interactions on the flat surface, the geometric constraint of the dynamics is also an attractive topic when modeling organs such as the brain. This paper reports pattern dynamics in a neural field equation defined on spheroids as model curved surfaces. We treat spot solutions as localized patterns and discuss how the geometric properties of the curved surface change their properties. To analyze spot patterns on spheroids with small flattening, we first construct exact stationary spot solutions on the spherical surface and reveal their stability. We then extend the analysis to show the existence and stability of stationary spot solutions in the spheroidal case. One of our theoretical results is the derivation of a stability criterion for stationary spot solutions localized at poles on oblate spheroids. The criterion determines whether a spot solution remains at a pole or moves away. Finally, we conduct numerical simulations to discuss the dynamics of spot solutions with the insight of our theoretical predictions. Our results show that the dynamics of spot solutions depend on the curved surface and the coordination of neural interactions.

This paper presents a hybrid modeling approach that couples an Agent-Based Model (ABM) with a partial differential equation (PDE) model in an epidemic setting to simulate the spatial spread of infectious diseases using a compartmental structure with seven health states. The goal is to reduce the computational complexity of a full-ABM by introducing a coupled ABM-PDE model that offers significantly faster simulations while maintaining comparable accuracy. Our results demonstrate that the hybrid model not only reduces the overall simulation runtime (defined as the number of runs required for stable results multiplied by the duration of a single run) but also achieves smaller errors across both 25% and 100% population samples. The coupling mechanism ensures consistency at the model interface: agents crossing from the ABM into the PDE domain are removed and represented as density contributions, while surplus density in the PDE domain is used to generate agents with plausible trajectories derived from mobile phone data. We evaluate the hybrid model using real-world mobility and infection data for the Berlin-Brandenburg region in Germany, showing that it captures the core epidemiological dynamics while enabling efficient large-scale simulations.

Early brain development is crucial for lifelong neurodevelopmental health. However, current clinical practice offers limited knowledge of normal embryonic brain anatomy on ultrasound, despite the brain undergoing rapid changes within the time-span of days. To provide detailed insights into normal brain development and identify deviations, we created the 4D Human Embryonic Brain Atlas using a deep learning-based approach for groupwise registration and spatiotemporal atlas generation. Our method introduced a time-dependent initial atlas and penalized deviations from it, ensuring age-specific anatomy was maintained throughout rapid development. The atlas was generated and validated using 831 3D ultrasound images from 402 subjects in the Rotterdam Periconceptional Cohort, acquired between gestational weeks 8 and 12. We evaluated the effectiveness of our approach with an ablation study, which demonstrated that incorporating a time-dependent initial atlas and penalization produced anatomically accurate results. In contrast, omitting these adaptations led to anatomically incorrect atlas. Visual comparisons with an existing ex-vivo embryo atlas further confirmed the anatomical accuracy of our atlas. In conclusion, the proposed method successfully captures the rapid anotomical development of the embryonic brain. The resulting 4D Human Embryonic Brain Atlas provides a unique insights into this crucial early life period and holds the potential for improving the detection, prevention, and treatment of prenatal neurodevelopmental disorders.

A rational framework is proposed to explain how we accommodate unbounded sensory input within bounded memory. According to this framework, memory is stored as a statistic-like representation that is repeatedly summarized and compressed to make room for new input. Summarization of sensory input must be rapid; that of abstract trace might be slower and more deliberative, drawing on elaborative processes some of which might occasionally reach consciousness (as in mind-wandering). Short-term sensory traces are summarized as simple statistics organized into structures such as a time series, graph or dictionary, and longer-term abstract traces as more complex statistic-like structures. Summarization at multiple time scales requires an intensive process of memory curation which might account for the high metabolic consumption of the brain at rest. Summarization may be guided by heuristics to help choose which statistics to apply at each step, so that the trace is useful for a wide range of future needs, the objective being to "represent the past" rather than tune for a specific task. However, the choice of statistics (or of heuristics to guide that choice) is a potential target for learning, possibly over long-term scales of development or evolution. The framework is intended as an aid to make sense of our extensive empirical and theoretical knowledge of memory and bring us closer to understanding it in functional and mechanistic terms.

Neural activity fluctuates over a wide range of timescales within and across brain areas. Experimental observations suggest that diverse neural timescales reflect information in dynamic environments. However, how timescales are defined and measured from brain recordings vary across the literature. Moreover, these observations do not specify the mechanisms underlying timescale variations, nor whether specific timescales are necessary for neural computation and brain function. Here, we synthesize three directions where computational approaches can distill the broad set of empirical observations into quantitative and testable theories: We review (i) how different data analysis methods quantify timescales across distinct behavioral states and recording modalities, (ii) how biophysical models provide mechanistic explanations for the emergence of diverse timescales, and (iii) how task-performing networks and machine learning models uncover the functional relevance of neural timescales. This integrative computational perspective thus complements experimental investigations, providing a holistic view on how neural timescales reflect the relationship between brain structure, dynamics, and behavior.

Background: Any treatment of multiple sclerosis should preserve mental function, considering how cognitive deterioration interferes with quality of life. However, mental assessment is still realized with neuro-psychological tests without monitoring cognition on neuro-biological grounds whereas the ongoing neural activity is readily observable and readable. Objective: The proposed method deciphers electrical brain states which as multi-dimensional cognetoms quantitatively discriminate normal from pathological patterns in an EEG. Method: Baseline recordings from a prior EEG study of 88 subjects, 36 with MS, were analyzed. Spectral bands served to compute cognetoms and categorize subsequent feature combination sets. Result: The brain states predictor correlates with disease burden and duration. Using cognetoms and spectral bands, a cross-sectional comparison separated patients from controls with a precision of 85% while using bands alone arrived at 79%. Conclusion: We demonstrate the efficiency of the quantitative data-driven method based on brain states analysis by contrasting EEG data of patients with MS and healthy subjects. The congruity with disease severity and duration is a neurophysiological indicator for disease accumulation over time. We discuss potential applications of the approach for the monitoring of disease time course and treatment efficacy in longitudinal clinical studies in psychiatry and neurology.

Language models, especially transformer-based ones, have achieved colossal success in NLP. To be precise, studies like BERT for NLU and works like GPT-3 for NLG are very important. If we consider DNA sequences as a text written with an alphabet of four letters representing the nucleotides, they are similar in structure to natural languages. This similarity has led to the development of discriminative language models such as DNABert in the field of DNA-related bioinformatics. To our knowledge, however, the generative side of the coin is still largely unexplored. Therefore, we have focused on the development of an autoregressive generative language model such as GPT-3 for DNA sequences. Since working with whole DNA sequences is challenging without extensive computational resources, we decided to conduct our study on a smaller scale and focus on nucleotide sequences of human genes rather than the whole DNA. This decision has not changed the structure of the problem, as both DNA and genes can be considered as 1D sequences consisting of four different nucleotides without losing much information and without oversimplification. Firstly, we systematically studied an almost entirely unexplored problem and observed that RNNs perform best, while simple techniques such as N-grams are also promising. Another beneficial point was learning how to work with generative models on languages we do not understand, unlike natural languages. The importance of using real-world tasks beyond classical metrics such as perplexity was noted. In addition, we examined whether the data-hungry nature of these models can be altered by selecting a language with minimal vocabulary size, four due to four different types of nucleotides. The reason for reviewing this was that choosing such a language might make the problem easier. However, in this study, we found that this did not change the amount of data required very much.

Many biochemical measurements define state spaces that are bounded, absorbing, and physically irreversible, yet are routinely analysed using linear and Euclidean frameworks that assume global invertibility, symmetry, and translation invariance. This mismatch can irretrievably obscure biological structure, independent of data quality, scale, or preprocessing. This work formalises the structure of bounded biochemical state spaces using cysteine redox regulation as a representative example and identify the minimal algebraic properties required for categorically correct representation. Hard boundaries, absorbing states, and irreversible ensemble dynamics render linear algebra incompatible with these objects. This work demonstrates that tropical algebra provides a natural realisation of the required properties by replacing additive linear structure with order-based, piecewise-linear operations that encode dominance, saturation, and path dependence without contradiction. By making non-invertibility and absorption explicit rather than implicit, this framework resolves a fundamental algebraic mismatch and establishes a principled foundation for the representation and analysis of bounded biochemical data.

Across neuroscience, artificial intelligence, and related fields, dominant models of intelligence typically privilege convergence: uncertainty is reduced, competing explanations are eliminated, and behaviour is governed by the optimisation of a single objective or policy. While this framing has proved powerful in many settings, it sits uneasily with biological and adaptive systems that maintain redundancy, ambiguity, and parallel explanatory processes over extended timescales. Here we propose an alternative perspective, termed polyphonic intelligence, in which coherent behaviour and meaning emerge from the coordination of multiple semi-independent inferential processes operating under shared constraints. Rather than resolving plurality through dominance or collapse, polyphonic systems sustain multiple explanatory trajectories and integrate them through soft alignment, compatibility relations, and bounded influence. We develop this perspective conceptually and formally, introducing a variational framework in which multiple coordinated approximations are maintained without winner-takes-all selection. This formulation makes explicit how plurality can remain stable, tractable, and productive, and clarifies how polyphonic inference differs from ensemble methods, mixture models, and Bayesian model averaging. Through proof-of-principle examples, we demonstrate that non-dominating, pluralistic inference can be implemented in simple computational systems without requiring centralised control or global convergence. We conclude by discussing implications for neuroscience, psychiatry, and artificial intelligence, and by arguing that intelligence may be more fruitfully understood as coordination without command rather than as the elimination of uncertainty.

Biological neural networks self-organize according to local synaptic modifications to produce stable computations. How modifications at the synaptic level give rise to such computations at the network level remains an open question. Pehlevan et al. [Neur. Comp. 27 (2015), 1461--1495] proposed a model of a self-organizing neural network with Hebbian and anti-Hebbian synaptic updates that implements an algorithm for principal subspace analysis; however, global stability of the nonlinear synaptic dynamics has not been established. Here, for the case that the feedforward and recurrent weights evolve at the same timescale, we prove global stability of the continuum limit of the synaptic dynamics and show that the dynamics evolve in two phases. In the first phase, the synaptic weights converge to an invariant manifold where the `neural filters' are orthonormal. In the second phase, the synaptic dynamics follow the gradient flow of a non-convex potential function whose minima correspond to neural filters that span the principal subspace of the input data.

Sickle cell disease causes erythrocytes to become sickle-shaped, affecting their movement in the bloodstream and reducing oxygen delivery. It has a high global prevalence and places a significant burden on healthcare systems, especially in resource-limited regions. Automated classification of sickle cells in blood images is crucial, allowing the specialist to reduce the effort required and avoid errors when quantifying the deformed cells and assessing the severity of a crisis. Recent studies have proposed various erythrocyte representation and classification methods. Since classification depends solely on cell shape, a suitable approach models erythrocytes as closed planar curves in shape space. This approach employs elastic distances between shapes, which are invariant under rotations, translations, scaling, and reparameterizations, ensuring consistent distance measurements regardless of the curves' position, starting point, or traversal speed. While previous methods exploiting shape space distances had achieved high accuracy, we refined the model by considering the geometric characteristics of healthy and sickled erythrocytes. Our method proposes (1) to employ a fixed parameterization based on the major axis of each cell to compute distances and (2) to align each cell with two templates using this parameterization before computing distances. Aligning shapes to templates before distance computation, a concept successfully applied in areas such as molecular dynamics, and using a fixed parameterization, instead of minimizing distances across all possible parameterizations, simplifies calculations. This strategy achieves 96.03\% accuracy rate in both supervised classification and unsupervised clustering. Our method ensures efficient erythrocyte classification, maintaining or improving accuracy over shape space models while significantly reducing computational costs.

We establish a generalized work theorem for stochastic chemical reaction networks (CRNs). By using a compensated Poisson jump process, we identify a martingale structure in a generalized entropy defined relative to an auxiliary backward process and extend nonequilibrium work relations to processes stopped at bounded arbitrary times. Our results apply to discrete, mesoscopic chemical reaction networks and remain valid for singular initial conditions and state-dependent termination events. We show how martingale properties emerge directly from the structure of reaction propensities without assuming detailed balance. Stochastic simulations of a simple chemical kinetic proofreading network are used to explore the dependence of the exponentiated entropy production on initial conditions and model parameters, validating our new work theorem relationships. Our results provide new quantitative tools for analyzing biological circuits ranging from metabolic to gene regulation pathways.