Blood-based biomarkers underpin clinical diagnosis and management, yet their interpretation relies largely on fixed population reference intervals that ignore stable, intra-patient variability. As such, population-based interpretation can mask meaningful deviation from an individual's baseline, risking delayed disease detection. To remedy this, there have been increasing efforts to personalize blood biomarker interpretation using individual testing histories. However, these methods may overfit to sparse data, inflating false-positive rates and unnecessary follow-up, and can also unwittingly include unrecognized or subclinical disease. Here, we leverage nearly 2 billion longitudinal laboratory measurements from over 1.6 million individuals across North America, the Middle East, and East Asia, to show that while laboratory values are highly individual, purely personalized intervals routinely overfit, classifying up to 68% of measurements as abnormal, without corresponding associations with adverse clinical outcomes. We then introduce NORMA, a conditional transformer-based framework that generates reference intervals by conditioning on both a patient's history and population-level data about "normal" variation. NORMA-derived intervals achieve higher precision for predicting outcomes, including mortality, acute kidney injury, and chronic disease. These findings caution against over-personalization in laboratory medicine and demonstrate that anchoring individual trajectories to population-level priors outperforms either approach alone. To promote transparency, we publicly release the model, code, and an interactive user interface for accessible, individualized laboratory interpretation.

Spatial patterns arising from the collective behavior of individual agents are present across biological systems. While agent-based models offer a natural framework for uncovering unknown agent (e.g., cell) interactions, these stochastic models face significant challenges. For spatial patterns, agent-based modeling often involves manual tuning to attain qualitative consistency with multiple experiments. This process limits predictive power and raises questions about parameter identifiability and model uniqueness. Combining topological techniques and Bayesian computation, we present a multi-objective methodology for parameter inference in detailed models. We illustrate our approach by inferring parameters in an agent-based model of zebrafish patterns, achieving practical identifiability in several case studies. By introducing extended prior distributions, we then reframe parameter inference as rule inference, allowing us to search across over 80 candidate agent-based rules to identify an alternative, simpler model consistent with our data.

We present the Open-Source Sleep Monitor and Modulator (OSSMM), an open-source hardware and software platform for accessible sleep research. The OSSMM comprises a small wearable headband built from 3D prints and affordable commercial-off-the-shelf (COTS) components at a material cost under 40 euros, supported by a companion Android application. The system requires no conductive gels, disposable electrodes, or specialized equipment, and captures multiple biosignals movement, pulse, electrooculography (EOG), and putative electroencephalography (EEG) with wireless connectivity for data storage and potential sleep modulation capability via an onboard vibration motor. A proof-of-concept single-participant evaluation across 15 nights demonstrated that the captured biosignals support four-stage sleep classification (Wake, Light Sleep, Deep Sleep, REM) using conventional machine learning methods, with the best-performing model achieving a Macro F1-score of 0.770 and accuracy of 0.776 against a validated non-contact sleep monitor ($κ$=0.63 with PSG). Two technical findings are of particular note. First, inexpensive, reusable conductive thermoplastic polyurethane (CTPU) electrodes from commercial fitness chest straps captured a differential signal whose spectral properties in canonical EEG frequency bands, including signatures consistent with sleep spindles, are the principal features driving classification. Second, this signal is obtained from just two frontal electrodes without a dedicated ground reference, suggesting that practical sleep staging is achievable with simpler configurations than typically employed. All hardware designs, software, and build instructions are openly available to support replication and modification by the research community.

Incorporating vaccination into mathematical models appears deceptively simple: models integrate vaccine-derived protections, such as reduced susceptibility to infection, using parameters informed by empirical estimates of vaccine efficacy or effectiveness (VE). In practice, however, empirical VE estimates often do not correspond directly to the parameters of epidemiological models. Here, we extend previous work to demonstrate that in order to accurately parameterize a model, one must consider both a vaccine's mechanism of action and the statistic used to infer VE from empirical data. When a vaccine confers leaky protection -- that is, vaccination partially rather than completely reduces individual infection risk -- we show that common empirical VE estimation methods do not provide directly applicable values for model parameters. Naive (i.e. direct) incorporation of these VE estimates into models results in an underestimate of population-level vaccine impact. To make progress when these estimates are the only available sources for VE, we introduce a parameterization approach which more accurately aligns the modeled effect of vaccination with empirical estimates. Under this adjusted parameterization approach, models predict fewer total infections and lower herd immunity thresholds for leaky vaccines than would be predicted under current parameterization practices. Our parameterization guidelines and adjustment approach can be used to improve accuracy in models that are used in vaccine decision making and public health planning.

The brain learns abstract representations of high-dimensional sensory input, but the plasticity rules that enable such learning are unknown. We study biologically plausible algorithms on the Random Hierarchy Model (RHM), an artificial dataset designed to investigate how deep neural networks learn the intrinsic hierarchical structure of high-dimensional data. We focus on two types of local learning rules that avoid both a long convergence time and the use of a symmetric error network. The first type uses direct feedback signals to approximate error propagation from the output layer. The second type uses layerwise self-supervised contrastive or non-contrastive loss functions that do not explicitly approximate errors at the output layer. We show that all rules of the first type fail to solve the tasks of the RHM and trace this failure back to input-specific nonlinearities (`masking') that are implemented in full backpropagation and are essential for learning complex tasks. However, algorithms of the second type are able to learn the hierarchical hidden structure of the RHM tasks and are as data-efficient as supervised backpropagation training, while being compatible with known rules of synaptic plasticity in cortex.

Classifying protein topology is essential for deciphering biological function, but progress is held back by the lack of large-scale benchmarks that avoid duplicates and by models that do not scale well. We introduce TEDBench, a large-scale, non-redundant benchmark for protein fold classification constructed from the Encyclopedia of Domains (TED) and Foldseek-clustered AlphaFold structures. We show that on TEDBench, current protein representation learning methods either require very large models or fail to deliver strong performance. To address this challenge, we propose Masked Invariant Autoencoders (MiAE), a self-supervised framework for protein structure representation learning. MiAE uses an extremely high masking ratio of up to 90% with an $\mathrm{SE(3)}$-invariant encoder and a lightweight decoder that reconstructs backbone coordinates from the latent representation and mask tokens. MiAE scales well and outperforms supervised counterparts and state-of-the-art baselines on TEDBench, establishing a strong recipe for protein fold classification. To test transfer beyond AlphaFold structures, we further benchmark on a curated dataset from experimental structures of CATH v4.4. TEDBench is available at https://github.com/BorgwardtLab/TEDBench.

Self-initiated attention shifts play a critical role in voluntary behavior but are difficult to study due to the absence of explicit temporal markers. While previous studies have examined their neural correlates, it remains unclear how multi-dimensional electroencephalography (EEG) features contribute to their characterization within an interpretable computational framework. In this study, we build on an experimental paradigm developed in our previous work, which enables controlled comparison between task-constrained self-initiated shifts and externally instructed shifts under identical visual stimulation. Within this setting, we investigate whether preparatory EEG activity can distinguish these two types of attention shifts. We adopt a machine learning-based approach and conduct two complementary analyses: (1) a performance-oriented assessment of frequency-specific topographic patterns, and (2) a model-based feature attribution analysis using SHapley Additive exPlanations (SHAP). These analyses provide a structured view of how spectral features across regions of interest contribute to model behavior. Our results demonstrate reliable within-subject classification performance, indicating that preparatory EEG activity contains subject-specific discriminative information within this paradigm. The analysis shows that higher-frequency bands and frontal regions contribute strongly to model decisions, although such contributions should be interpreted cautiously due to the potential influence of non-neural artifacts in high-frequency EEG signals. Overall, this work highlights the value of interpretable machine learning for analyzing subject-specific EEG signal patterns in a controlled experimental setting, with potential applications in personalized and asynchronous brain-machine interface systems.

In 2019, Paulsson and collaborators conjectured that stochastic biochemical control networks have fundamental limits on how much intrinsic noise can be simultaneously suppressed across multiple components. Ripsman, Kell, and Hilfinger recently proposed a formal proof strategy for unit-birth models based on a stationary information-theoretic decomposition. Here, we provide a rigorous mathematical justification for this argument. We consider continuous-time Markov chains on $\Z^N_{\ge 0}$ in which each component is degraded linearly and produced in unit births at a state-dependent rate depending on the other components but not on itself. Noise in component $i$ is measured by the Fano factor $F_{X_i}$, the ratio of stationary variance to mean, with Poisson value $1$ as baseline. Our first contribution is to isolate explicit hypotheses on moments, mean birth rates, and total-rate growth under which the formal information-flow identities can be rigorously justified. Following the proof outline of Ripsman, Kell, and Hilfinger, we then prove the conjecture. Our second contribution is to make these hypotheses checkable: a uniform positive lower bound on the birth rates and at-most-linear total growth dominated by the weakest degradation rate suffices, via Foster--Lyapunov methods. Our third contribution is a structural strengthening. Under a signed monotonicity condition on the rate functions, satisfied by structurally balanced signed interaction networks, we prove that the stationary distribution is associated with respect to the corresponding signed partial order. This upgrades the global tradeoff to the termwise bound $F_{X_i}\ge 1$ for every $i$. Hence, within the signed-monotone subclass, sub-Poissonian noise requires a frustrated interaction topology.

Videomicroscopy, when combined with machine learning, offers a promising approach for studying the early development of in vitro produced (IVP) embryos. However, manually annotating developmental events, and more specifically cell divisions, is time-consuming for a biologist and cannot scale up for practical applications. We aim to automatically classify the cell stages of embryos from 2D time-lapse microscopy videos with a deep learning approach. We focus on the analysis of bovine embryonic development using video microscopy, as we are primarily interested in the application of cattle breeding, and we have created a Bovine Embryos Cell Stages (ECS) dataset. The challenges are three-fold: (1) low-quality images and bovine dark cells that make the identification of cell stages difficult, (2) class ambiguity at the boundaries of developmental stages, and (3) imbalanced data distribution. To address these challenges, we introduce CLEmbryo, a novel method that leverages supervised contrastive learning combined with focal loss for training, and the lightweight 3D neural network CSN-50 as an encoder. We also show that our method generalizes well. CLEmbryo outperforms state-of-the-art methods on both our Bovine ECS dataset and the publicly available NYU Mouse Embryos dataset.

Contemporary computational neuroscience features two prominent modeling traditions. Bottom-up whole-brain modeling (WBM) builds biophysically detailed simulations of brain structure and dynamics, whereas top-down neuroconnectionism optimizes deep neural networks for functional performance. Each has achieved remarkable success yet remains incomplete with WBMs lacking functional competence and neuroconnectionist models showing limited biological grounding. Here we propose functional whole-brain models (fWBMs) as a unified modeling paradigm that integrates structural and dynamical realism with task-performing capacity. fWBMs are defined by four minimal criteria: structural grounding in empirical connectomes and regional biology, continuous-time dynamical realism, functional competence across cognitive domains, and mappable observables to neuroimaging, electrophysiologcal and behavioral data. To formalize this integration, we establish a three-pillar roadmap across short-, mid-, and long-term horizons, and outline the scientific and clinical opportunities this paradigm enables. We argue that the disciplined pursuit of this integrative vision will generate the tools, common language, and cross-scale hypotheses needed to advance our understanding of the brain.

Dengue fever poses a persistent public health challenge in rapidly urbanizing Indian cities such as Ahmedabad, where spatial heterogeneity and seasonal variability complicate forecasting and control. In this study, we develop a data-driven compartmental framework to simulate transmission dynamics, generate forecasts, and evaluate intervention strategies across the Ahmedabad Municipal Corporation (AMC). We employ a Mechanistic Seasonal Dengue Transmission (M-SDT) model that incorporates symptomatic and asymptomatic infections. We calibrated the proposed model using zone-wise dengue case data during 2020--2024. Parameter uncertainty is rigorously quantified using a bootstrap sampling framework with negative binomial noise. The calibrated model reveals pronounced spatial heterogeneity across AMC zones, with persistent hotspots and distinct transmission regimes. Forecasts for 2026--2028 indicate continued endemic circulation with moderate inter-annual variability. Sensitivity analysis identifies the mosquito biting rate and vector mortality as dominant drivers of long-term disease burden, highlighting the central role of vector ecology in shaping epidemic outcomes. Evaluating seasonal vector control strategies shows a notable difference in operation; periodic fogging has a cumulative effect over the years, while sustained residual spraying can quickly curb outbreaks and decrease incidence by over 80%. The zone-wise analysis reveals that the mosquito-to-human ratio governs not only the baseline outbreak potential but also each zone's responsiveness to control strategies. Overall, the M-SDT modelling framework enables reconstruction of unobserved dynamics, rigorous uncertainty quantification, and evaluation of targeted, zone-specific interventions, underscoring the importance of integrating fine-scale surveillance data with mechanistic modelling for adaptive urban dengue control.

AlphaFold3 introduces a diffusion-based architecture that elevates protein structure prediction to all-atom resolution with improved accuracy. This state-of-the-art performance has established AlphaFold3 as a foundation model for diverse generation and design tasks. However, its iterative design substantially increases inference time, limiting practical deployment in downstream settings such as virtual screening and protein design. We propose DCFold, a single-step generative model that attains AlphaFold3-level accuracy. Our Dual Consistency training framework, which incorporates a novel Temporal Geodesic Matching (TGM) scheduler, enables DCFold to achieve a 15x acceleration in inference while maintaining predictive fidelity. We validate its effectiveness across both structure prediction and binder design benchmarks.

Despite similarities between models exhibiting absorbing phase transitions (APTs) and those showing Kardar-Parisi-Zhang (KPZ) growth, the relationship between these universal fluctuations has remained elusive. We numerically study (1+1)-dimensional interfaces of (2+1)-dimensional models showing APTs of directed percolation (DP) and compact directed percolation (CDP) classes with an active boundary, finding a universal crossover from short-time APT-governed fluctuations to long-time KPZ fluctuations. Upon rescaling time and length by the APT correlation time and length, the cumulants of the interface height distributions collapse onto a single scaling function. The fluctuation properties of the discrete Domany-Kinzel model and the continuum stochastic Fisher-Kolmogorov-Petrovsky-Piskunov (sFKPP) equation coincide, indicating that the KPZ growth parameters are determined solely by fundamental properties of the APT. For the CDP sFKPP equation, a dimensionless parameter tunes both the critical interface distribution and the KPZ parameters, with the interface properties of the biased voter model recovered in a limiting case. These results uncover a universal crossover in which KPZ fluctuations emerge from APT fluctuations at long times, linking paradigmatic universality classes of nonequilibrium scale-invariant phenomena.

Large-scale hypothesis testing is central to modern science, where controlling the False Discovery Rate (FDR) has become the standard approach to managing false positives across many simultaneous tests. Hypotheses rarely exist in isolation; they often exhibit structure through proximity, connectivity, or hierarchy. This structure represents both a challenge and an opportunity: while classical methods treat these dependencies as obstacles requiring conservative correction, leveraging them can substantially increase discovery power. Here, we reframe structured FDR control as a regularized learning problem. By optimizing within a suitable Reproducing Kernel Hilbert Space (RKHS), we introduce a framework that unifies continuous domains, graphs, and hierarchies under a single algorithm through kernel choice alone. This formulation enables smooth solutions in place of the piecewise-constant fits of prior methods, principled likelihood-based hyperparameter selection rather than heuristic tuning, and inference at unobserved locations which in turn supports sample-efficient experimental design. Building on this estimator, we provide two decision rules which we prove to control the FDR. We validate our method on two sources: spatial locations derived from high-dimensional real-world datasets, and a differential gene expression task utilizing protein-protein interaction graphs.

Von Economo neurons (VENs) are selectively lost in behavioural-variant frontotemporal dementia (bvFTD) and reduced in autism spectrum conditions (ASC), yet their computational role in social learning remains unexplained. We train a spiking neural network (the VENCircuit) embedding VEN-like projection neurons (K=40, 2% of total) in a recurrent pyramidal circuit across 50 matched random initialisations with and without VENs. The network is trained on a controlled binary classification task; we make no claim to model social cognition directly. VEN-intact networks converged in 49/50 cases (98%) versus 35/50 (70%) for VEN-ablated networks (Fisher's exact OR=21.0, 95% CI 2.7-167, p=8.7e-5). Failed ablated networks showed complete absence of learning, inconsistent with a speed-of-learning account. Phase-ablation experiments show VEN removal is most disruptive during mid-training (epochs 5-25), when a co-adaptive dependency forms in the pyramidal circuit. We derive a formal account showing VENs provide a direct gradient pathway immune to Jacobian instabilities affecting the recurrent circuit. Inference-time VEN ablation caused a significant performance drop (Wilcoxon p=0.022), ranging from no change (16/20 networks) to catastrophic collapse (0.989 to 0.620). VENs function as acquisition scaffolds whose developmental absence produces stochastic learning failure - a computational analogue of variable social skill acquisition in ASC - with falsifiable predictions for organoid and electrophysiology studies.

Understanding what individual neurons encode is a core question in neuroscience. In primary visual cortex (V1), mathematical models (e.g., Gabor functions) capture neural selectivity, but no comparable framework exists for higher areas. We show that natural language can fill this role: across macaque V1 and V4, the selectivity of most neurons is captured by concise, verifiable semantic descriptions. Using digital twins of V1 and V4, we develop a closed-loop framework that translates each neuron's high- and low-activating images into dense captions, generates a semantic hypothesis and synthesized images, and verifies the hypothesis in silico. Descriptions range from oriented edges and spatial frequency in V1 to conjunctions of form, color, and texture in V4. In V4, images generated from activating and suppressing hypotheses drove 96.1% of neurons above the 95th and 97.6% below the 5th percentile of natural-image responses, respectively (vs. ~10% for random images); V1 activation results matched V4, while V1 suppression was less describable in language. Representational similarity analysis reveals partial alignment between neural activity, vision embeddings, and language embeddings, with vision most aligned to neural activity; alignment lost in the text bottleneck is recovered when hypotheses are rendered back into images, showing that linguistic compression is lossy yet semantically faithful. Together, these results show that combining generative models with neural digital twins enables interpretable, testable descriptions of neural function at scale, toward agentic scientific discovery.

Protein design aims to compose amino-acid sequences that fold into stable three-dimensional structures while satisfying targeted functional properties. The field is increasingly shifting toward vibe protein design, where a single model is expected to generate novel sequences, engineer existing proteins, and reason about protein characteristics through flexible natural-language constraints. Large language models (LLMs) have emerged as a leading paradigm in this space. However, existing evaluation benchmarks often limit their scope to a partial aspect of protein design, while others restrict design objectives to structured input schemas, lacking an integrated framework that evaluates the broad spectrum of protein design competence under open-ended intents. To this end, we present Vibe Protein design Benchmark (VibeProteinBench), a language-interfaced benchmark that probes generalist capabilities through three complementary stages mirroring a computational protein design workflow: recognition, engineering, and generation. Each stage is grounded in expert-curated mechanistic rationales and multi-faceted in silico validation, to computationally verify whether model outputs are biologically plausible. Evaluations across diverse general-purpose and domain-specialized LLMs reveal that no model achieves strong performance across all three stages, suggesting that generalist protein design remains a substantial open challenge for current LLMs.

The identification of reliable molecular biomarkers for Parkinson's disease remains challenging due to its multifactorial nature. Although protein sequences constitute a fundamental and widely available source of biological information, their standalone discriminative capacity for complex disease classification remains unclear. In this work, we present a controlled and leakage-free evaluation of multiple representations derived exclusively from protein primary sequences, including amino acid composition, k-mers, physicochemical descriptors, hybrid representations, and embeddings from protein language models, all assessed under a nested stratified cross-validation framework to ensure unbiased performance estimation. The best-performing configuration (ProtBERT + MLP) achieves an F1-score of 0.704 +/- 0.028 and ROC-AUC of 0.748 +/- 0.047, indicating only moderate discriminative performance. Classical representations such as k-mers reach comparable F1 values (up to approximately 0.667), but exhibit highly imbalanced behavior, with recall close to 0.98 and precision around 0.50, reflecting a strong bias toward positive predictions. Across representations, performance differences remain within a narrow range (F1 between 0.60 and 0.70), while unsupervised analyses reveal no intrinsic structure aligned with class labels, and statistical testing (Friedman test, p = 0.1749) does not indicate significant differences across models. These results demonstrate substantial overlap between classes and indicate that primary sequence information alone provides limited discriminative power for Parkinson's disease classification. This work establishes a reproducible baseline and provides empirical evidence that more informative biological features, such as structural, functional, or interaction-based descriptors, are required for robust disease modeling.

Temporal point processes offer a powerful framework for sampling from discrete distributions, yet they remain underutilized in existing literature. We show how to construct, for any target multivariate count distribution with downward-closed support, a multivariate temporal point process whose event-count vector in a fixed-length sliding window converges in distribution to the target as time tends to infinity. Structured as a system of potentially coupled infinite-server queues with deterministic service times, the sampler exhibits a discrete form of momentum that suppresses random-walk behaviour. The admissible families of processes permit both reversible and non-reversible dynamics. As an application, we derive a recurrent stochastic neural network whose dynamics implement sampling-based computation and exhibit some biologically plausible features, including relative refractory periods and oscillations. The introduction of auxiliary randomness reduces the sampler to a birth-death process, establishing the latter as a degenerate case with the same limiting distribution. In simulations on 63 target distributions, our sampler always outperforms these birth-death processes and frequently outperforms Zanella processes in multivariate effective sample size, with further gains when normalized by CPU time.

During music listening, cortical activity encodes both acoustic and expectation-related information. Prior work has shown that ANN representations resemble cortical representations and can serve as supervisory signals for EEG recognition. Here we show that distinguishing acoustic and expectation-related ANN representations as teacher targets improves EEG-based music identification. Models pretrained to predict either representation outperform non-pretrained baselines, and combining them yields complementary gains that exceed strong seed ensembles formed by varying random initializations. These findings show that teacher representation type shapes downstream performance and that representation learning can be guided by neural encoding. This work points toward advances in predictive music cognition and neural decoding. Our expectation representation, computed directly from raw signals without manual labels, reflects predictive structure beyond onset or pitch, enabling investigation of multilayer predictive encoding across diverse stimuli. Its scalability to large, diverse datasets further suggests potential for developing general-purpose EEG models grounded in cortical encoding principles.

The boundaries of cooperative helix--coil transitions directly affect protein allostery and conformational dynamics, yet the physical origin of the persistent one-to-two-residue assignment ambiguity at these structural interfaces remains unresolved. We apply the discrete Hasimoto map to translate three-dimensional protein backbone geometry into a one-dimensional discrete nonlinear Schrödinger effective potential and analyze its spatial-frequency fluctuations. Helical segments display near-integrable, low-entropy soliton-like states, while coil regions exhibit broadband conformational noise. Statistical analysis of over 19,000 boundaries across 1,986 proteins reveals a median geometric transition width of only 0.145 residues, providing an independent kinematic counterpart to the high thermodynamic cooperativity of the Zimm--Bragg model. This sub-residue spatial narrowness indicates an intrinsic observational constraint governed by the Gabor uncertainty principle, whereby any macroscopic spectral probe tends to blur the microscopic phase boundary, suggesting that the boundary ambiguity in structural biology is not merely algorithmic but reflects a physical resolution limit inherent to the biopolymer lattice.

Spatial transcriptomics provides an unprecedented perspective for deciphering tissue spatial heterogeneity. However, high-resolution spatial transcriptomic technology remains constrained by limited gene coverage, technical complexity, and high cost. Existing spatial transcriptomics super-resolution methods from low resolution data suffer from two fundamental limitations: poor out-of-distribution generalization stemming from a neglect of inherent biological heterogeneity, and a lack of physical consistency. To address these challenges, we propose SRast, a novel physically constrained generalist framework designed for robust spatial transcriptomics super-resolution. To tackle heterogeneity, SRast employs a strategic decoupling architecture that explicitly decouples gene semantics representation from spatial geometry deconvolution, utilizing self-supervised learning to align latent distributions and mitigate cross-sample shifts. Regarding physical priors, SRast reformulates the task as ratio prediction on the simplex, performing a flow matching model to learn optimal transport-based geometric transformations that strictly enforce local mass conservation. Extensive experiments across diverse species, tissues, and platforms demonstrate that SRast achieves state-of-the-art performance, exhibiting superior zero-shot generalization capabilities and ensuring physical consistency in recovering fine-grained biological structures.

We introduce a nonparametric model for inferring time-evolving, unobserved probability distributions from discrete-time data consisting of unlabelled partitions. The latent process is a two-parameter Poisson-Dirichlet diffusion, and observations arise via exchangeable sampling. Applications include social and genetic data where only aggregate clustering summaries are observed. To address the intractable likelihood, we develop a tractable inferential framework that avoids label enumeration and direct simulation of the latent state. We exploit a duality between the diffusion and a pure-death process on partitions, together with coagulation operators that encode the effect of new data. These yield closed-form, recursive updates for forward and backward inference. We compute exact posterior distributions of the latent state at arbitrary times and predictive distributions of future or interpolated partitions. This enables online and offline inference and forecasting with full uncertainty quantification, bypassing MCMC and sequential Monte Carlo. Compared to particle filtering, our method achieves higher accuracy, lower variance, and substantial computational gains. We illustrate the methodology with synthetic experiments and a social network application, recovering interpretable patterns in time-varying heterozygosity.

The design of genome-scale constraint-based metabolic networks has steadily advanced, with an increasing number of successful cases achieving growth-coupled production, in which the biosynthesis of key metabolites is linked to cell growth. However, a major cause of design failures is the inability to find solutions within realistic time limits. Therefore, it is essential to develop methods that achieve a high success rate within the specified computation time. In this study, we propose a framework for ranking the importance of individual genes to accelerate the solution of the original mixed-integer linear programming (MILP) problems in the design of constraint-based models. In the proposed method, after pre-assigning values to highly important genes, the MILPs are solved in parallel as a series of mutually exclusive subproblems. It is found that our framework was able to recover most of the successful cases identified by the original approach and achieved a 37% to 186% increase in success rate compared to the original method within the same time limits. Analysis of the MILP solution process revealed that the proposed method reduced the sizes of subproblems and decreased the number of nodes in the branch-and-bound tree. This framework for ranking gene importance can be directly applicable to a range of MILP-based algorithms for the design of constraint-based metabolic networks. The developed scripts are available on \href{https://github.com/MetNetComp/Gene-Ranked-RatGene}{https://github.com/MetNetComp/Gene-Ranked-RatGene}.

Purpose: To develop a deep learning method for the automatic segmentation of spinal nerve rootlets on various MRI scans. Material and Methods: This retrospective study included MRI scans from two open-access and one private dataset, consisting of 3D isotropic 3T TSE T2-weighted (T2w) and 7T MP2RAGE (T1-weighted [T1w] INV1 and INV2, and UNIT1) MRI scans. A deep learning model, RootletSeg, was developed to segment C2-T1 dorsal and ventral spinal rootlets. Training was performed on 76 scans and testing on 17 scans. The Dice score was used to compare the model performance with an existing open-source method. Spinal levels derived from RootletSeg segmentations were compared with vertebral levels defined by intervertebral discs using Bland-Altman analysis. Results: The RootletSeg model developed on 93 MRI scans from 50 healthy adults (mean age, 28.70 years $\pm$ 6.53 [SD]; 28 [56%] males, 22 [44%] females) achieved a mean $\pm$ SD Dice score of 0.67 $\pm$ 0.09 for T1w-INV2, 0.65 $\pm$ 0.11 for UNIT1, 0.64 $\pm$ 0.08 for T2w, and 0.62 $\pm$ 0.10 for T1w-INV1 contrasts. Spinal-vertebral level correspondence showed a progressively increasing rostrocaudal shift, with Bland-Altman bias ranging from 0.00 to 8.15 mm (median difference between level midpoints). Conclusion: RootletSeg accurately segmented C2-T1 spinal rootlets across MRI contrasts, enabling the determination of spinal levels directly from MRI scans. The method is open-source and can be used for a variety of downstream analyses, including lesion classification, neuromodulation therapy, and functional MRI group analysis.

Gene expression analysis holds the key to many biomedical discoveries, yet extracting insights from raw transcriptomic data remains formidable due to the complexity of multiple large, semi-structured files and the need for extensive domain expertise. Current automation approaches are often limited by either inflexible workflows that break down in edge cases or by fully autonomous agents that lack the necessary precision for rigorous scientific inquiry. GenoMAS charts a different course by presenting a team of LLM-based scientists that integrates the reliability of structured workflows with the adaptability of autonomous agents. GenoMAS orchestrates six specialized LLM agents through typed message-passing protocols, each contributing complementary strengths to a shared analytic canvas. At the heart of GenoMAS lies a guided-planning framework: programming agents unfold high-level task guidelines into Action Units and, at each juncture, elect to advance, revise, bypass, or backtrack, thereby maintaining logical coherence while bending gracefully to the idiosyncrasies of genomic data. On the GenoTEX benchmark, GenoMAS reaches a Composite Similarity Correlation of 89.13% for data preprocessing and an F$_1$ of 60.48% for gene identification, surpassing the best prior art by 10.61% and 16.85% respectively. Beyond metrics, GenoMAS surfaces biologically plausible gene-phenotype associations corroborated by the literature, all while adjusting for latent confounders. Code is available at https://github.com/Liu-Hy/GenoMAS.

In single-cell research, tracing and analyzing high-throughput single-cell differentiation trajectories is crucial for understanding biological processes. Key to this is the robust modeling of hierarchical structures that govern cellular development. Traditional methods face limitations in computational cost, performance, and stability. VAE-based approaches have made strides but still require branch-specific network modules, limiting their scalability and stability, while often suffering from posterior collapse. To overcome these challenges, we introduce HDTree, a generative modeling framework designed for robust lineage inference. HDTree captures tree relationships within a hierarchical latent space using a unified hierarchical codebook and employs a quantized diffusion process to model continuous cell state transitions. By aligning the generative process with the Waddington landscape, this method not only improves stability and scalability but also enhances the biological plausibility of inferred lineages. HDTree's effectiveness is demonstrated through comparisons on both general-purpose and single-cell datasets, where it outperforms existing methods in lineage inference accuracy, reconstruction quality, and hierarchical consistency. These contributions enable accurate and efficient modeling of cellular differentiation paths, offering reliable insights for biological discovery.\footnote{Code is available at https://github.com/zangzelin/code\_HDTree\_icml.

Colorectal cancer (CRC) poses a major public health challenge due to its increasing prevalence, particularly among younger populations. Microsatellite instability-high (MSI-H) CRC and deficient mismatch repair (dMMR) CRC constitute 15% of all CRC and exhibit remarkable responsiveness to immunotherapy, especially with PD-1 inhibitors. Despite this, there is a significant need to optimise immunotherapeutic regimens to maximise clinical efficacy and patient quality of life. To address this, we employ a novel framework driven by delay integro-differential equations to model the interactions among cancer cells, immune cells, and immune checkpoints in locally advanced MSI-H/dMMR CRC (laMCRC). Several of these components are being modelled deterministically for the first time in cancer, paving the way for a deeper understanding of the complex underlying immune dynamics. We consider two compartments$\unicode{x2014}$the tumour site and the tumour-draining lymph node (TDLN)$\unicode{x2014}$taking into account phenomena such as DC migration, T cell proliferation, and CD8+ T cell exhaustion and reinvigoration. Parameter values and initial conditions are derived from experimental data, integrating various pharmacokinetic, bioanalytical, and radiographic studies, along with deconvolution of bulk RNA-sequencing data from the TCGA COADREAD and GSE26571 datasets. We finally optimised neoadjuvant treatment with pembrolizumab, a widely used PD-1 inhibitor, to balance efficacy, efficiency, and toxicity in laMCRC patients. We mechanistically analysed factors influencing treatment success and improved upon currently FDA-approved therapeutic regimens for metastatic MSI-H/dMMR CRC, demonstrating that a single medium-to-high dose of pembrolizumab may be sufficient for effective tumour eradication while being efficient, safe, and practical.

We consider an SLIARS mathematical epidemiology model including intervention in the form of vaccination and treatment. Contrary to classical models, it is assumed that treatment doses can be limited in availability. Mathematically, we show that most scenarios actually reduce to classic well-known scenarios: having an unreplenished number of doses is akin to having none, while being able to restore stocks is (often) equivalent to the classic situation with vaccination and treatment. We also perform a computational analysis, illustrating some of the transient and stochastic dynamics that diverge from deterministic long-term behaviour, as well as the impact of budgetary constraints.

Ecological communities are often characterized by many weak and few strong interspecific interactions, yet their quantitative structure, generative basis, and links to community-level properties remain poorly understood. Using two empirical datasets of plant--animal networks, we show that both trophic and mutualistic interaction strengths distribute skewed weak and Pareto-strong tails (SWAPS), as quantified by positive skewness and extreme value theory, respectively. We further find that interaction strengths are taxon-specific and largely constrained within taxa. In community assembly simulations based on a generalized Lotka--Volterra model, this taxonomic conservatism, together with multiple interaction types beyond trophic and mutualistic ones, is required for the emergence of SWAPS distribution. Notably, SWAPS distribution emerges not only at the species level but also across lineages, and its emergence accompanies increases in community diversity and complexity. Together, these results identify SWAPS distribution as a previously unrecognized interaction signature of ecological communities and provide a new perspective on the organization of community-level properties.