Nearby neurons in cortex share similar response profiles, producing systematic spatial organization across sensory and cognitive systems. Recent topographic models reproduce aspects of this structure but remain unimodal and spatially constrain each layer separately, yielding fragmented maps that capture neither the contiguity of cortical processing streams nor their integration across modalities. We introduce Topo-Omni, a topographic multimodal model in which visual, auditory, and language/cognitive processing share a single contiguous in-silico sheet. Built by fine-tuning a pretrained foundation model with a spatial smoothness objective, this architecture develops clusters across modalities that are consistent with human neuroimaging, from sensory to cognitive systems. Driving or suppressing a cluster selectively biases or impairs perception, paralleling human intervention studies. Finally, we use our model to screen for novel clusters in-silico and discover new natural landscape and animal networks which we validate in human data. A single spatial principle thus organizes representations across modalities and processing stages, yielding testable hypotheses about cortical organization.

Spatially resolved transcriptomics (SRT) is transforming how we study tissues by measuring gene expression in cells in their spatial context. However, the field lacks robust methodological guidance on one of its most fundamental analytical steps: how to accurately segment cells and assign spatially localized transcripts to them. Major technical challenges include sparse molecular signals, transcript displacement, complex cellular morphologies, and the projection of three-dimensional tissue architecture onto two-dimensional imaging planes. These challenges make segmentation a major source of uncertainty, with errors that can propagate through downstream analyses and ultimately lead to misleading biological interpretations. Here, we argue that segmentation should be treated as a central unresolved problem in spatial omics rather than a routine preprocessing step. We review current approaches, highlight key methodological limitations, including the lack of appropriate metrics and gold-standard benchmarks, and propose a community-driven path forward. Establishing shared evaluation frameworks, scalable benchmark datasets, and transparent reporting standards will be essential for transforming SRT into a robust and reproducible foundation for biological discovery and clinical translation.

Ask a pretrained biomedical language model whether "cortisol 28 ug/dL" and "stock-market volatility" are related, and it returns a cosine similarity of 0.83 on a scale where 1.0 means identical. The two share no mechanism. This is not a corner case: every off-the-shelf biomedical encoder we tested (BioBERT, PubMedBERT, BioM-ELECTRA) scores unrelated cross-domain pairs between 0.76 and 0.92 when the answer should be near zero. Accuracy on cross-domain discrimination is 0%. Retrieval systems survive this, because a language model downstream filters the noise. A Large Behavioural Model (LBM), a foundation model whose subject is a person rather than a sentence, does not: it reasons over a graph of a user's life and treats embedding proximity as evidence that two events are causally linked. False proximity writes a false causal edge, and everything downstream inherits the error. Here, embedding geometry is not a tuning knob; it is correctness. We report the fix. A contrastive pass over 72,034 pairs raises PubMedBERT BIOSSES correlation from 0.633 to 0.828 and within-vs-across-domain separation from 1.05x to 1.63x. A second pass, BODHI, mines hard negatives from edges absent in a biomedical knowledge graph and lifts separation to 2.30x and the discrimination gap to +0.392, at a 4.5% BIOSSES cost. On an Intel Xeon 6737P with AMX, OpenVINO cuts single-query latency from 1367 ms to 10 ms (133x) and reaches 555 sentences/sec. One finding contradicts standard advice: FP16 beats INT8 on this silicon at every serving batch size, and we explain why. The same model on a no-AMX Ice Lake instance runs 13-27x slower. We release the benchmark suite, training corpora, the BODHI generator, and the OpenVINO scripts.

Motivation: Transformer-based models are increasingly applied to large-scale single-cell transcriptomics, showing strong performance through self-supervised learning on millions of cells. However, most existing approaches treat genes as independent features, and largely ignore prior biological knowledge, which limits interpretability and robustness. In this paper, we explore whether explicitly incorporating gene regulatory information can improve both model performance and biological insight. Results: We present scTransformer, the first Transformer-based approach that builds a priori knowledge of biological mechanisms into the model's attention patterns. By constraining information flow according to known regulatory structures, the model learns representations that are more biologically meaningful. We evaluate scTransformer on a disease-relevant single-nucleus RNA-seq dataset using supervised cell-type classification. Compared to standard Transformers, our approach improves classification accuracy, enhances separation of cell types in embedding space, and produces attention patterns consistent with known regulatory programs. Overall, our results demonstrate that embedding biological structure into Transformer models can enhance interpretability without sacrificing performance, offering a principled step toward biologically grounded foundation models for single-cell omics.

Ecological communities with structured interactions exhibit collective phenomena such as percolation and clustering of occupied sites. While these effects have been documented in experiments and simulations, systematic analytical understanding has remained limited. In this paper, we develop a dynamical theory of these phenomena for competitive ecological systems defined on random interaction graphs. We introduce a discrete version of the generalized Lotka-Volterra model that preserves key macroscopic features of continuous ecological dynamics while enabling analytical treatment. Within this framework, we characterize the emergence of percolating clusters and describe the spatial organization of surviving sites. Our analysis uncovers which equilibria can be reached by the dynamics and shows how this dynamical accessibility governs the onset of clustering and percolation. In doing so, our framework complements classical Lotka-Volterra theory by providing a dynamical perspective on the collective organization of structured communities.

Living organisms, from bacteria to humans, are more likely to survive if their traits enhance fitness. In populations well adapted to their environmental niches, natural selection proceeds via rarely beneficial mutations. But when a catastrophe wipes out niche diversity, sudden adaptation often follows. Here, we present a data-validated theory of natural selection in the wake of catastrophe and unveil a simple law that emerges during recovery: the mean fitness relaxes inversely with time, with a prefactor proportional to the number of traits coupled to the post-catastrophe environment. We put our approach to test using experimental fitness landscapes measured following antibiotic administration to E. coli. The resulting mean trait adaptation is not described by gradient ascent on a fitness landscape, instead it follows an algorithm known as Levenberg-Marquardt optimization. Near fitness peaks, evolutionary trajectories are biased against greediness - from an optimization perspective, post-catastrophic selection is optimistic.

Fundamental investigations into how different molecular encoding methods affect molecular property prediction remain relatively limited. In this study, we extensively examined the optimal molecular encoding methods for molecular properties prediction using two prevalent structure designs: a classical neural network model (MLP) and a Transformer encoder-based model (MLP+TL). For molecular encoding methods, we investigated several types of fingerprints, including traditional topological fingerprints, substructure-based fingerprints, and string-based representations. These two models were trained on seven well-known molecular datasets to evaluate different input molecular encoding methods based on evaluation metrics. On several biologically relevant classification tasks, including toxicity, mutagenicity, and side-effect prediction, our models consistently achieved average AUC values above 0.9. Rather than relying on external post-hoc explanation methods such as the local interpretable model-agnostic explanation (LIME) or the Deep SHapley Additive exPlanations (SHAP), we leveraged the model's intrinsic attention weights as an internal interpretability signal for identifying potentially important feature. The MLP+TL model using MACCS and PubChem as input can capture chemically interpretable groups that determined the major blood-brain barrier (BBB) permeability and mutagenicity in Salmonella typhimurium. In particular, a comparison between Morphine and Heroin highlighted the role of hydroxyl-related substructures in BBB permeability prediction, which was consistently reflected in the attention weights. Overall, our findings provide practical guidance for selecting effective molecular encoding methods and contribute to the development of interpretable molecular informatics approaches for drug discovery.

Spinal pathology is a leading cause of pain and disability worldwide. Spine MRI is central to clinical evaluation, yet its interpretation remains complex and time-consuming, requiring integration of information across multiple imaging sequences and anatomical regions. Despite recent advances in automated MRI analysis, effectively combining multi-sequence data while preserving sequence-specific diagnostic information remains an open challenge. Here we present SpineAgent, a multi-agent framework for spine MRI report generation built upon a multi-sequence foundation model trained on routine clinical data from 32,047 patients and 453,683 MRI series, comprising a total of 13,441,191 MRI slices. To accommodate diverse modalities of sequences, we first pre-train two DINOv3-based encoders separately on T1- and T2-weighted sequences. We then introduce a continual training strategy that learns a synthesizer to embed images of other sequences using the T1 and T2 encoders, producing patient-level embedding that integrates various signals across MRI sequences. Using these embeddings, SpineAgent achieves state-of-the-art performance, and demonstrates strong generalizability under cross-manufacturer and cross-cohort evaluation. Beyond classification, SpineAgent enables pathology localization by identifying findings-relevant slices and segmenting pathological regions. It also supports multimodal image-report retrieval, providing a solid foundation for scalable and explainable MRI report generation. We further integrate these validated capabilities of SpineAgent into 37 specialized agents. Finally, we incorporate their outputs as structured tokens within a Medical Report Agent trained end-to-end for report generation. Through both automated metrics and expert evaluation by five radiologists, SpineAgent achieves leading performance in spine MRI report generation.

When do three-dimensional conformer ensembles improve molecular property prediction beyond two-dimensional fingerprints? We provide the first systematic, mechanistically grounded answer. Through ~1,000 experiments spanning 13 model configurations, 14 regression targets, and 2 classification targets across MoleculeNet, QM9, and MARCEL benchmarks, we discover selective complementarity: conformer ensemble statistics extracted via Distribution Kernel Operators (DKOs) yield statistically significant RMSE reductions on solvation-dependent properties (ESOL -11.0%, p < 10^{-9}; FreeSolv -13.5%, p < 3x10^{-5}; 10-seed paired validation) while providing no benefit for electronic or steric tasks. Three lines of evidence confirm this selectivity has a physical rather than statistical basis: improvement is larger under scaffold splits than random splits (+11.9% vs. +8.5% on ESOL), concentrates on large, flexible molecules (+18.9% for heaviest quartile), and grows monotonically with training data. We establish a four-tier performance hierarchy: end-to-end 3D GNNs (SchNet, PaiNN; 21-42% over fingerprints) >= engineered physicochemical descriptors (PMI/SASA/USR) > Morgan fingerprints + XGBoost > all neural conformer ensemble methods, confirmed by two architecturally diverse GNNs and revealing that the pre-computed feature bottleneck limits ensemble approaches. Feature attribution and mutual information analysis expose the mechanistic asymmetry: conformer mean features carry 2-8x more information per feature than fingerprint bits, yet covariance features contribute <2% of model signal, explaining why five simple scalar invariants outperform all complex covariance architectures (p < 0.001). These findings yield an empirical property taxonomy and a practical decision framework for when conformer generation is worth the investment.

Dynamical mean-field theory recasts deterministic chaos in random recurrent networks as an effective stochastic process. We show that for analytic nonlinearities with sufficiently fast Fourier decay, this stochasticity is only apparent: the continuous past of a realized mean-field trajectory uniquely determines its future. The mean-field theory is therefore not merely an ensemble description, but a conditional prediction theory for individual trajectories. Unfolding the power spectrum into a Krylov state space exposes how this latent determinism is organized across an infinite hierarchy of temporal modes. The associated Krylov growth rate sets the complexity of finite-resolution prediction and upper-bounds the largest Lyapunov exponent in this class of networks. Thus, microscopic sensitivity and predictive complexity are distinct aspects of mean-field chaos. Our results extend Krylov growth ideas developed for Hamiltonian chaotic dynamics to classical dissipative systems.

A sufficient account of how the neocortex learns must meet three criteria: 1. Computationally, it must approximate a powerful, general-purpose learning algorithm known to scale to human-level intelligence; 2. Algorithmically, it must be implementable using known, well-established neural circuits within the neocortex and associated brain structures; 3. Implementationally, there must be a detailed account for how all of the algorithmic mechanisms actually function at a neurochemical level. At present, there is only one framework that meets all of these criteria: error-driven predictive learning via temporal derivatives, driven by corticothalamic circuits, based on competitive kinase synaptic plasticity induction mechanisms. This has been implemented in the Axon neural simulation framework using spiking neurons, and demonstrated to learn across a wide range of challenging cognitively motivated tasks.

The ability to predict protein three-dimensional structures from amino acid sequences is a landmark achievement in molecular biology, where recent deep learning approaches such as AlphaFold are the culmination of decades of work. Yet, the quantitative understanding of how protein sequences give rise to dynamic conformational changes and higher-order assemblies remains unsolved. Folding and conformational states are dynamic, stochastic processes, shaped by sequence, energy, co-translational constraints, chaperone machineries, and the physicochemical conditions of the cellular environment. Recent advances now position the field to move beyond static structural endpoints toward a mechanistic understanding of folding dynamics in living systems. Single-molecule techniques enable time-resolved observation of folding trajectories and intermediate states hitherto hidden by traditional structural biology approaches, while computational innovations and data-driven approaches offer new ways to integrate heterogeneous data across scales. In this Roadmap, we review the current conceptual landscape of protein folding, examine the experimental and theoretical gaps that remain, and discuss emerging strategies that integrate high-resolution measurements with multiscale modeling. We outline a roadmap toward a quantitative and predictive science of protein folding dynamics, conformational kinetics, and macromolecular self-assembly. Realizing this vision would transform our understanding of the dynamics of molecular self-organization, from the folding of individual polypeptides to the emergence of dynamic macromolecular complexes. This will enable rational control of folding and misfolding in health and disease, extend protein engineering principles beyond static structural design, and establish a mechanistic foundation for predictive and personalized interventions in proteostasis-related disorders.

Ordinary differential equation models are widely used to understand and forecast complex dynamical systems, but their predictive value depends on reliable parameter estimation. Structural identifiability assesses whether parameters can be uniquely recovered from ideal observations, whereas practical identifiability depends on finite, noisy and partially observed data. We introduce the Practical Identifiability Index (PII), a marginal uncertainty-width metric based on the logarithmic span of confidence intervals. Expressed on an order-of-magnitude scale, the PII summarises how tightly individual positive-valued parameters are constrained by available observations, enabling comparison across parameters, models, error structures and observation designs. The PII is intended as a complementary diagnostic, not a standalone identifiability test, and should be interpreted alongside coverage, profile likelihoods, posterior summaries, sensitivity analysis or structural identifiability results. Using parametric bootstrap experiments across growth and compartmental epidemic models, we identify consistent principles: uncertainty decreases as calibration windows become more informative, increases with observation noise and parameter coupling, and remains high for latent or indirectly observed processes. Parameters governing early observable dynamics become constrained sooner, while additional observables can improve constraint for latent progression and recovery parameters. The PII provides a simple, reportable summary of marginal parameter uncertainty for dynamical modelling.

Phylogenetic networks are graphs inferred from molecular sequence data that represent ancestral histories shaped by reticulate processes such as recombination, hybridization, and horizontal gene transfer. We introduce a family of distance metrics for rooted, ranked, unlabeled phylogenetic networks, extending a previously developed distance for ranked trees. Our approach relies on a bijective triangular matrix representation of phylogenetic networks that captures the temporal order of internal events, speciations, and hybridizations. Our metrics, defined as standard matrix norms, allow efficient quantitative comparisons of network topologies, timed networks and networks with differing numbers of hybridizations. Our distance can be used for both isochronous networks where all tips are sampled at one time point, and heterochronous networks where tips are allowed to be sampled at different time points. We show that our metrics capture biologically meaningful differences among evolutionary histories in both simulations and empirical posterior distributions of viral phylogenetic networks. These tools fill a methodological gap, enabling principled comparisons of ranked, unlabeled phylogenetic networks, including ancestral recombination graphs.

In April 2026, the MV Hondius expedition cruise ship became the site of the first documented cruise ship-associated Andes hantavirus (ANDV) cluster, with 13 confirmed and probable cases and 3 deaths among 149 passengers and crew. We applied a stochastic epidemic model to evaluate four embarkation scenarios under reproductive numbers anchored to published ANDV estimates. Scenario D, involving two latent infected persons at embarkation, was most consistent with the observed outbreak, yielding P(final size >= 13) = 11.6% and P(takeoff) = 58.5% at R0 = 2.12. Approximate Bayesian computation provided complementary support for multiple latent infections at embarkation, especially E1(0)=1 and E3(0)=2, but R0 remained weakly identifiable. A day-35 transmission reduction changed takeoff probability little in this counterfactual model. Findings support exposure-history assessment, early onboard surveillance, rapid isolation of symptomatic cases, and postdisembarkation monitoring for travelers from ANDV-endemic regions.

The Mader model is the most widely used mathematical framework for muscular energy metabolism in German-language sport science, underpinning lactate diagnostics, maximal lactate steady state (MLSS) estimation and training prescription. Despite decades of use, neither its dynamic ODE formulation nor its steady-state equations have been available as open code, leaving results based on the model impossible to reproduce independently. We close this gap with MetaboliSim, an open-source Python implementation of both formulations: a dynamic model that integrates the five-variable ODE system (phosphate potential, $\dot{V}\mathrm{O}_2$, muscle and blood lactate, and glycogen) with a fourth-order Runge-Kutta scheme, and a steady-state model that computes MLSS power and the lactate-power relationship in one- and two-compartment variants. We verified implementation correctness against published reference values and assessed physiological plausibility across constant-load, step-test, sprint and running protocols. The implementation reproduces the published reference output within stated tolerances and remains numerically stable throughout (halving the time step changes blood lactate by less than 0.01 mmol/L), with both formulations yielding congruent MLSS estimates. Key physiological behaviour ($\dot{V}\mathrm{O}_2$ on-kinetics, lactate accumulation, PCr dynamics and the sub/supra-MLSS separation) emerges directly from the model equations without protocol-specific tuning, and a sensitivity analysis shows MLSS power varying approximately linearly with $\dot{V}\mathrm{O}_{2\max}$ and nonlinearly with $\dot{V}\mathrm{La}_{\max}$. As the first openly available implementation of the complete Mader model (AGPL-3.0), MetaboliSim lets independent groups reproduce, verify and build on published model-based results. Source code: https://codeberg.org/3phos/metabolisim; Platform: https://metabolisim.org

Most statistical and machine learning methods for directed interactions focus on pairwise effects among variables. Even existing cyclic models represent feedback primarily through node-level dependencies, making large-scale recurrent organization difficult to estimate and compare. This limitation is particularly acute in biological and neural systems, where interactions are highly recurrent and involve many overlapping cycles. We introduce a variational framework for statistical inference on cyclic interactions. Directed interactions are represented as edge flows on a simplicial complex and evolved under an energy-minimizing dynamical system. The resulting dynamics separate transient interaction components from persistent harmonic flows, yielding a low-dimensional cycle space that captures stable recurrent organization. Rather than enumerating individual cycles, the proposed framework represents cyclic interactions as elements of a Hilbert space, enabling projection, averaging, comparison, and population-level statistical inference. We establish theoretical properties of the harmonic projection, including characterization of the cycle space, variance reduction, and population inference. Simulations demonstrate substantially improved recovery of cyclic structure in dense recurrent systems compared with existing directed-interaction methods. Applied to resting-state fMRI from 400 human subjects, the framework reveals reproducible large-scale cyclic organization that is not detectable through edgewise averaging. These results provide a scalable statistical framework for studying recurrent interactions in high-dimensional dynamical systems.

Token aggregation is a common bottleneck in models that map token representations to sample-level predictions, yet most pooling methods operate only in the original token domain. We propose FLaG, a plug-in aggregation module that transforms token representations with the real FFT, summarizes spectral components with learnable latent queries, applies a channel-wise gate, and reconstructs enhanced time-domain tokens for final pooling. We evaluate FLaG on antimicrobial peptide (AMP) activity prediction with ESM2, image classification with ResNet18 on CIFAR-10 and CIFAR-100, and text classification with RoBERTa on IMDB and GLUE. FLaG achieves its clearest gains on the ESM2-8M antimicrobial peptide tasks and on CIFAR-100, while remaining competitive with strong text baselines on IMDB and GLUE. Then we probe its behavior on the AMP setting with band knockouts, gate summaries, residue perturbations, latent-query readouts, and structure-proxy stratification. We find that low-frequency bands contribute the most overall, and the remaining higher-band pattern is more sample-specific. The gate acts as a broadly shared spectral reweighting stage and the cross-attention patterns are sample-specific with mild query-wise differentiation, and higher-helix peptides exhibit stronger average spectral sensitivity in both bacteria. The supplementary materials, source code and data are released at https://www.healthinformaticslab.org/supp/ and https://github.com/Kewei2023/AMPCliff/tree/FLaG.

DNA cis-regulatory elements (CREs) such as enhancers control gene expression levels. Accurately predicting regulatory activity from DNA sequences is valuable but challenging, as it requires understanding complex biological regulatory processes. Existing methods typically regress activity scores from sequences in a black-box manner, limiting both interpretability and regression performance. Meanwhile, large language models (LLMs) benefit from explicit reasoning processes, yet directly applying LLMs to raw DNA sequences performs poorly. In this paper, we bridge this gap by introducing R3LM, a framework that teaches LLMs reasoning-informed regression on regulatory DNA through structured biological knowledge. Specifically, we design a biologically grounded data format that structures DNA's regulatory information for improved LLM understanding, and construct CRE-ReasonBench, the first dataset that associates DNA sequences and activity scores with mechanistic reasoning traces. Through two-stage training that first teaches LLMs reasoning over structured biological information then performs regression, R3LM achieves state-of-the-art performance on enhancer prediction across three cell types, outperforming both LLMs with raw sequence input and specialized DNA models while providing interpretable mechanistic explanations. We expect R3LM as an interpretable reward model that can effectively assist biologists in CRE design. Code is available at https://github.com/DuanYi516/R3LM.

Eukaryotic DNA replication must remain robust under thermal, chemical, and genotoxic stress despite large fluctuations in replication dynamics. Here, we develop a lattice-based stochastic Monte Carlo framework for whole-genome replication in Saccharomyces cerevisiae at single base-pair resolution, incorporating probabilistic origin firing, replication fork-speed distributions, and a time-dependent limiting factor that governs the availability of cellular replication resources. The model is benchmarked quantitatively against experimental replication profiles before being applied to stress conditions, and reproduces diverse replication stress responses using only two effective parameters. Importantly, the analysis reveals that replication fork-speed heterogeneity underlies the emergence of Erlang-distributed S-phase durations and rare, anomalously prolonged replication events observed experimentally in Escherichia coli and human cell lines, while predicting similar behavior in S. cerevisiae. The framework further predicts non-monotonic thermal behavior, power-law scaling under hydroxyurea stress, and total replication-time dynamics under diverse genotoxic conditions.

This study analyses the Ako tidal flat in the Seto Inland Sea, Japan, where nearly all Zostera marina disappeared within a single year in 2025. Using aerial photographs from the 1940s onward, high-resolution satellite imagery, GRUS images (2.5-5 m), and monthly Sentinel-2 composites (10 m), we reconstructed approximately 80 years of seagrass distribution. YOLO-based segmentation using deep learning achieved high accuracy (overall accuracy >= 0.9) across these datasets; although species could not be discriminated, the models captured the major temporal dynamics in vegetation area. The long-term mean seagrass area was 6.8 ha, but values fluctuated widely, from 3.5 ha in 1974 to 41.3 ha in 1989 except 0.2 ha in 2025. Sentinel-2 composites from 2019 to 2026 revealed clear seasonality, with vegetation increasing in early summer and declining from autumn. In 2025, however, the area decreased sharply after summer and remained anomalously low throughout the winter of 2025-2026. Our results, indicating that the 2025 event was not a normal fluctuation but a rapid ecosystem shift involving the loss of the dominant canopy-forming species, most plausibly driven by regionally elevated summer water temperatures. The findings also have implications for seagrass Essential Ocean Variables (EOVs) and the State of Nature (SoN) metrics used in TNFD-aligned nature-related disclosures. Unlike forests, seagrass meadows require finer temporal resolution because both pronounced seasonality and abrupt collapse strongly influence area-based indicators. Therefore, in addition to previously noted issues such as species-level classification accuracy, we recommend that (1) baselines be defined over the longest available record and justified ecologically, (2) seasonal standardization be applied before inter-annual comparisons, and (3) years with extreme area anomalies be flagged rather than used as reference points.

Alzheimer's disease is a progressive neurodegenerative disorder, and its progression varies substantially across patients. Existing work aims to forecast patients' future cognitive state, with minimal focus on reconstructing the state from past visits. Furthermore, in current research, quantifying predictive uncertainty remains underexplored and relies on costly modalities such as MRI, PET, and CSF, limiting their deployment in resource-limited settings. In this research, our primary objectives are: First, bidirectional prediction of cognitive scores from irregular visits to present the complete disease trajectory. Second, to enable interpolation and extrapolation capabilities to assist clinicians in informed prognostic decision making, and third, to provide a well-calibrated uncertainty estimate for all predictions, and finally, to achieve the objectives using the modalities available during routine visits. We propose a unified framework, GNOVA: A GRU-Neural ODE Variational Autoencoder. The architecture combines a Gated Recurrent Unit encoder and a Neural ODE decoder within a variational autoencoder framework. In our work, we forecast the CDR-SB and MMSE Scores. The GRU encoder allows for any number of inputs at any time point. The Neural-ODE decoder performs continuous estimation, allowing interpolation and extrapolation at any desired time point. The Variational autoencoder allows for uncertainty estimation in predictions. We worked with 1,727 patients from the ADNI dataset over 10 years; the model achieved mean absolute errors of 1.35 and 2.28 for CDR-SB and MMSE scores, respectively, without requiring any neuroimaging or biomarker data. Feature-ablation studies revealed that age, BMI, and APOE4 status were strong predictors. The proposed framework enables the reconstruction of incomplete patient histories and the anticipation of future cognitive states.

Spatial transcriptomics (ST) is a powerful tool for exploring biological properties dependent on structure, proximity, and interaction in tissue. The methods underpinning ST are developing rapidly but are limited in their ability to profile many thousands of genes at a subcellular scale. Although dissociated from tissue, it is known that the whole-transcriptome readouts of cells in single-cell RNA sequencing (scRNA-seq) retain information about their former in situ neighbourhoods, motivating computational methods to recover it. While paired ST and scRNA-seq datasets are scarce, each modality in its own right is abundantly available. We therefore propose to perform cross-modal translation between unpaired ST and scRNA-seq data. In this work we show that a single-cell foundation model can perform this translation via adversarial fine-tuning. We demonstrate that our method performs favourably against methods built for multi-omics translation.

Objective: To develop and externally test a video-based framework for simultaneous detection of hyperkinetic MDs phenomenologies: dystonia, tremor, myoclonus, chorea, athetosis, ballismus, stereotypies, and tics using routine clinical recordings, with explicit testing of external, cross-cohort transfer from adult to pediatric populations. Methods: In this proof-of-concept study, the framework combines markerless pose estimation, kinematic descriptors, and a pretrained fondation model. A shared predictive backbone was developed on 21 adults with confirmed hyperkinetic MDs and 4 healthy controls assessed under a standardized protocol. External validation was performed on an independent external cohort: a real-world pediatric sample (n=12, monogenic combined MDs). For the external dataset, the backbone was deployed without retraining; lightweight calibration adjusted only the final subject-level decision step using a small labeled subset of patients selected by clinicians as representative of the cohort's phenotypic range. Results: After local calibration of the decision layer on the clinician-selected subset, performance improved consistently on the held-out pediatric patients (n=7): Hamming accuracy rose from 0.804 to 0.839 and the Jaccard index from 0.548 to 0.633. This calibrated performance was preserved, and the Jaccard index further improved, when the evaluation was restricted to the phenomenologies with more definite clinician agreement (Hamming accuracy 0.9, Jaccard index 0.786), indicating that the gains did not rest on the least-reliable labels.

Advances in machine learning and computational power have unlocked the predictive potential of the human genome, yet biologists now demand that these models also elucidate the underlying biological mechanisms. While interpretable machine learning (IML) techniques have been increasingly applied to bridge this gap, there has been a pervasive reliance on anecdotal validation: the vast majority of research relies on a single IML method and reports only isolated successful instances. Through a benchmarking study on transcription factor binding, we demonstrate the risks of current practices. We show that different IML methods can often (1) yield contradictory explanations for the same predictions, (2) fail to localize known regulatory motifs, and (3) fail to faithfully reflect the model's internal decision process. In light of this, we argue for a validation framework analogous to clinical trials: just as trials require rigorous design and adverse-event reporting, genomic interpretability must move beyond cherry-picked plausibility toward systematic assessment of consistency, faithfulness, and biological validity. To facilitate this, we propose a tiered framework to guide rigorous evaluation and reporting of genomic IML methods.

Protein function is largely determined by molecular surface geometry and physicochemical complementarity, yet most protein design methods condition only on backbone structure. We introduce SurfDesign, a surface-conditioned protein design framework that models molecular surfaces as continuous geometric manifolds and integrates them with pretrained protein language models. SurfDesign employs surface-based equivariant message passing to capture surface normals, curvature, and directional geometry, together with a parameter-efficient fine-tuning strategy. Focusing on functional protein design, we show that SurfDesign consistently outperforms prior surface-conditioned and backbone-only methods on de novo binder and enzyme design benchmarks. We also report strong performance on inverse-folding benchmarks as a diagnostic of structural compatibility. Our results highlight manifold-aware surface representations as a principled foundation for functional protein and enzyme design. Code is available at https://github.com/smiles724/SurfDesign.

RNA design consists of discovering a nucleotide sequence that optimizes predefined criteria, such as secondary structure. It is useful for synthetic biology, medicine, and nanotechnology. We propose Montparnasse, a Monte Carlo search framework based on Generalized Nested Rollout Policy Adaptation, augmented with a problem-specific prior, slow and long adaptation at level 1, and a lexicographic multicriteria evaluation. Montparnasse solves all 100 puzzles of the Eterna100 V1 benchmark consistently faster than DesiRNA, the previous state of the art, across all time limits, reaching full coverage more than three times faster overall. On messenger RNA secondary structure optimization for hemoglobin alpha, it identifies sequences with more paired bases than the MFE-optimal solution of LinearDesign.

Bulk gene expression profiling, which aggregates pooled RNA across cells within a biological sample, remains important in the single-cell era because it is typically less noisy, more sensitive, and more cost-effective than single-cell assays. Accordingly, a growing body of computational methods seeks to recover causal relations among genes from bulk expression data. However, aggregation is a lossy, non-invertible coarsening of the underlying cellular system, and it remains unclear whether and under what conditions causal relations are recoverable from aggregated bulk gene expression data. To answer this, we formalize recoverability under aggregation through two notions of consistency: functional-form consistency and conditional-independence consistency. We then derive necessary and sufficient conditions for recoverability, showing that these properties are preserved only under linear aggregations (e.g., sum/mean) coupled with affine structural equations. To assess the practical plausibility of these conditions, analyses of four bulk and four single-cell gene expression datasets further reveal that the estimated pairwise regulatory functions among genes deviate from linearity in both data types, providing limited empirical support for the linearity assumptions required for recoverability. Together, these results caution against recovering causal relations from aggregated bulk expression data without strong additional assumptions.

Across biological and social systems, cooperation often depends on phenotypic cues rather than random encounters. To account for real-world interactions unfolding across multiple, simultaneous dimensions, here we develop a general framework for the evolution of cooperation in multiplex networks governed by multi-phenotype homophily. We derive analytical conditions for natural selection to favor cooperation across phenotypic traits that are independent or exhibit epistasis and under different modes of mutation coupling. Despite the integration of fitness across layers, the conditions for cooperation resolve into layer-specific $σ$-rules, depending only on the local payoff structure, the effective number of phenotypes, and the mutation rates. We show that phenotypic diversity fosters cooperation by partitioning populations into assortative niches. Furthermore, in finite populations, intensifying the prisoner's dilemma shifts the dependence of cooperation on strategy mutation from monotonically decreasing, through U-shaped, to monotonically increasing. Our work provides a unified account of how multi-phenotype homophily underpins the evolutionary dynamics of cooperation in heterogeneous populations.

A long standing challenge in computational chemistry and biophysics is efficiently sampling the Boltzmann distribution of molecules. Advances in generative modeling have been proposed to address the limitations of conventional sampling techniques by eliminating the computational cost of simulation. A promising direction is iteratively finetuning diffusion models along a temperature ladder whereby training data is generated via importance sampling during inference-time annealing. Unfortunately, these methods require computing a divergence over the score field to estimate importance weights, rendering them intractable for larger systems. Here we present scalable inference-time annealing (SITA), which retrains flow-based models to generate samples at progressively lower temperatures using an energy-based model to facilitate fast surrogate likelihoods. We demonstrate state-of-the-art performance on both Alanine Dipeptide and Alanine Tripeptide while avoiding costly divergence terms. Our code is available at https://github.com/countrsignal/sita.git