Designing protein sequences that bind specific ligands benefits from an inverse-folding model conditioned on full ligand geometry. We present UMA-Inverse, which replaces the sparse graph backbone of LigandMPNN with a dense pair-representation encoder: a six-block PairMixer (triangle multiplication, no triangle self-attention or sequence track) refines all residue-residue and residue-ligand atom pairs, supervised by an auxiliary distogram objective, and an autoregressive decoder attends over ligand atoms through a learned, position-specific readout of the pair tensor. The model is compact ($\sim$3.3 M parameters). On the LigandMPNN test splits it reaches 56.1%/55.1%/35.3% interface recovery (small-molecule/metal/nucleotide). It trails LigandMPNN, but by less than the published numbers suggest: re-run under our identical protocol, LigandMPNN scores 59.8/64.4/53.3 (vs. published 63.3/77.5/50.5). In a pocket-fixed setting the redesigns are confidently folded and ligand-binding-competent under Boltz-2 cofolding, again modestly behind LigandMPNN. Its distinctive property is representational: the dense encoder propagates ligand identity to residues far beyond the interface, where LigandMPNN's signal decays. We offer UMA-Inverse as a compact baseline for ligand-conditioned inverse folding that trails LigandMPNN in accuracy, together with a characterization of how a dense all-pairs encoder distributes ligand information.

Disturbance regimes and nutrient inputs are changing worldwide, with consequences for the structure and functioning of plant communities. Classical life-history theory predicts that disturbance should shift communities from long-lived perennials toward short-lived annuals, and that nutrient enrichment may amplify this shift. However, these predictions have not been tested experimentally across broad environmental gradients. Here, using a global coordinated grassland experiment spanning 37 sites, we tested how physical disturbance, vegetation removal and shallow soil tillage, and fertilisation reshape annual-perennial balance, and whether disturbance relaxes the climatic limits of annual dominance. Disturbance nearly doubled the proportion of annual species and more than doubled the relative cover of annuals, whereas fertilisation had little influence and did not interact with disturbance. The disturbance-driven shift arose through contrasting pathways: in graminoids and legumes, it reflected the loss of perennial cover, while in forbs, the expansion of annual cover. In the absence of disturbance, annual dominance was restricted to systems with extremely hot and dry summers, but disturbance nearly tripled the extent of climate space in which annuals dominated. By rapidly reassembling after disturbance, annuals may help maintain vegetation cover, but their expansion also signals loss of perennial cover and the long-term ecosystem functions associated with it.

We propose a phenomenological model of the Global Neuronal Workspace (GNW) in which early sensory processing generates an effective complex-valued landscape governing the dynamics of high-level stimulus representations. This landscape provides a dynamical bridge between sensory encoding and conscious access, enabling both processes to be described within a unified framework. High-level representations are encoded in a cloud function defined on a Hilbert space over a perceptual state space, thereby combining the holistic structure of mental images with a neural implementation. Its dynamics is governed by a nonlinear Schrödinger-type equation in imaginary time with a non-Hermitian, non-normal Hamiltonian and a nonlinear Lotka--Volterra-type term that preserves norm and enables spatially nonlocal interactions. The Hermitian and anti-Hermitian parts of the Hamiltonian generate complementary processes: recognition via dissipative localization at minima of the GNW landscape and information broadcasting via spatial spreading across the state space. The resulting dynamics reproduces the subliminal--preconscious--conscious hierarchy of sensory processing. Conscious access corresponds to the emergence of a bound state, which occurs only when both the GNW landscape depth and the degree of top-down attention exceed threshold values. The resulting framework provides a tractable dynamical description linking sensory processing, attention, and conscious access within a unified dynamical setting.

Respiratory syncytial virus (RSV) is a leading cause of bronchiolitis and other lower respiratory tract infections in infants. Increased viral circulation in the post-COVID era and heterogeneous prevention strategies across regions have made RSV control more challenging. We develop a stage-structured, age-stratified Susceptible-Infected-Recovered (SIR) compartmental model tailored to the Italian setting to investigate the population-level impact of infant prophylaxis with Nirsevimab, a long-acting monoclonal antibody. Scenario-based simulations over a multi-year horizon show that increasing infant protection coverage substantially reduces RSV incidence among infants and also yields indirect benefits in older age groups. In particular, extending coverage to infants born outside the epidemic season further lowers cumulative incidence, although infant-targeted prophylaxis alone does not reduce the control reproduction number below the epidemic threshold in the parameter range explored. These findings suggest that broader and more consistent infant Nirsevimab coverage may reduce RSV burden and support the evaluation of alternative implementation strategies in the Italian context.

Current computational approaches for drug design typically focus on generating molecules conditioned on specific targets or general molecular properties, often neglecting the influence of disease context on target behavior and therapeutic outcomes. To address this gap, we introduce DrugGen-2, a novel generative model that designs small molecules conditioned on both disease ontology and target protein sequences. DrugGen-2 was developed by fine-tuning a pre-trained GPT-2 model on a curated dataset of approved drugs linked to their diseases and targets, using a two-step strategy of supervised fine-tuning followed by reinforcement learning via group relative policy optimization (GRPO). This process was guided by reward functions optimizing for chemical validity, novelty, diversity, and high predicted binding affinity. When evaluated on five protein targets relevant to diabetic nephropathy, DrugGen-2 significantly outperformed baseline models (DrugGPT and DrugGen). It demonstrated a superior capacity to generate unique molecules, exhibited greater structural similarity to approved drugs, and achieved improved predicted binding affinities across all targets. Molecular docking analyses further supported these findings, identifying candidate ligands with strong binding potential, including compounds with predicted affinities (-9.917, -9.485, and -9.367) exceeding those of reference drugs such as enalapril for angiotensin-converting enzyme (-8.283). By integrating disease-specific context into molecular generation, DrugGen-2 advances AI-assisted drug discovery, offering a powerful tool for de novo design and drug repurposing that accounts for the complex interplay between diseases and molecular targets.

Genomic prediction models often fail to transfer across institutions because sequencing panels differ across sites, creating structural feature missingness at deployment. Existing approaches to this challenge typically restrict analysis to genes shared across cohorts, exclude patients with incomplete profiles, or rely on test-time imputation, all of which can reduce robustness and limit the use of multi-center data. We propose Survival prediction Handling Incomplete Features using Transformer (SHIFT), a missingness-aware survival model that directly predicts from incomplete genomic inputs without test-time imputation. SHIFT represents each genomic feature separately and uses masked self-attention, along with a feature-availability mask, so that predictions are based only on observed inputs. Further, we introduce variable-rate feature masking during training to improve robustness to heterogeneous missingness patterns. We evaluate the approach on glioblastoma and lung squamous cell carcinoma with external validation across multiple cohorts, including a challenging setting with severe cross-cohort panel mismatch. Across these settings, SHIFT shows strong generalization and compares favorably with standard survival baselines and imputation-based approaches, while using a single model across differing feature sets. We also find that incorporating patients from incomplete cohorts during development can improve performance on external data, suggesting that partially observed cohorts need not be excluded from model building. These results support missingness-aware modeling as a practical strategy for multi-center survival prediction in precision oncology.

Energy Matching has emerged as a powerful generative framework that combines flow model efficiency with the explicit likelihood of Energy-Based Models (EBMs) via a single, time-independent scalar potential. However, directly training this potential on high-dimensional 3D data remains computationally challenging. While distilling a pre-trained flow model circumvents some of the initial training costs, we demonstrate that velocity fields inevitably contain non-conservative rotational artifacts (curl). Forcing a strictly conservative scalar potential to match this unconstrained field creates a "structural conflict", which degrades generation quality and mode coverage. To solve this, we propose Projected Energy Matching, a scalable framework that resolves these structural and computational bottlenecks. We introduce Helmholtz Distillation, a structural relaxation that leverages a Hutchinson trace estimator to explicitly absorb rotational noise into an auxiliary residual network. We subsequently refine this landscape using Negative Caching, a memory-efficient strategy that reuses negative samples across micro-batches, rendering sampling tractable during contrastive training with gradient accumulation. We deploy our method as an unconditional prior for real-world medical CT inverse problems, specifically sparse-view reconstruction. Ultimately, our amortized pipeline reduces total compute to a small fraction of that required by standard energy matching, while achieving high-fidelity reconstructions and successfully resolving severe measurement artifacts.

A series of results from the NeuroAI over the past fifteen years have raised core questions both about how to compare Deep Neural Network (DNN) models to the brain, and about how much convergent evolution to expect between artificial networks and real brain networks. Here, we show that for any two minimal DNN solutions to a sufficiently hard task: (i) "weak" alignment of network representations based on affine mappings guarantees "strong" alignment of privileged axes, and (ii) alignment "zippers" up the network hierarchy, causing the emergence of privileged axes from end-to-end task optimization. These results formalize the notion of contravariance from Cao and Yamins [2024], and illustrate important consequences for the theory of NeuroAI: with sufficiently strong tasks, choice of metric for inter-network comparison is not all that sensitive, and that convergent evolution is probably inevitable.

Gene drive alleles bias their own inheritance to offspring. They can fix in a wild-type population in spite of a fitness cost, and even lead to the eradication of the target population if the fitness cost is high. However, this outcome may be prevented or delayed if areas previously cleared by the drive are recolonised by wild-type individuals. Here, we investigate the conditions under which these stochastic wild-type recolonisation events are likely and when they are unlikely to occur in one spatial dimension. More precisely, we examine the conditions ensuring that the last individual carrying a wild-type allele is surrounded by a large enough number of drive homozygous individuals, resulting in a very low chance of wild-type recolonisation. To do so, we make a deterministic approximation of the distribution of drive alleles within the wave, and we split the distribution of wild-type alleles into a deterministic part and a stochastic part. Our analytical and numerical results suggest that the probability of wild-type recolonisation events increases with lower fitness of drive individuals and with smaller local carrying capacity. Numerical simulations show that these results extend to two spatial dimensions. The role of the migration rate however, is less clear but has a lower impact. We further demonstrate that, in the event of wild-type recolonization, the probability of subsequent drive reinvasion decreases with smaller values of the intrinsic growth rate of the population. Overall, our study paves the way for further analysis of wild-type recolonisation at the back of eradication traveling waves.

In a 2018 paper and a subsequent article published in 2023, researchers reported that mitochondria maintain temperatures 10oC-15oC higher than the surrounding cytoplasm-a finding that deviates by five to six orders of magnitude from theoretical predictions based on Fourier's law of heat conduction. In 2022, we proposed a solution to this apparent paradox. In the present perspective, we build upon that framework and introduce new ideas to further unravel how a biological membrane-whether of an organelle or a whole cell-can become significantly warmer than its environment. We propose that ion-translocating proteins embedded in the inner mitochondrial membrane (IMM) can be modeled as ratchet engines, introducing a novel, previously overlooked mode of heat transfer. This mechanism, coupled with localized heat release during the cyclical dehydration-translocation-hydration of ions through membrane proteins, may generate transient but substantial temperature spikes. The cumulative thermal occupancy of these microscopic events across the three-dimensional surface of the IMM can account for the elevated temperatures detected by molecular probes.

High-dimensional neural activity often reside in a low-dimensional subspace, referred to as neural manifolds. Grid cells in the medial entorhinal cortex provide a periodic spatial code that are organized near a toroidal manifold, independent of the spatial environment. Due to the periodic nature of its code, it is unclear how the brain utilizes the toroidal manifold to understand its state in a spatial environment. We introduce a novel framework that decodes spatial information from grid cell activity using topology. Our approach uses topological data analysis to extract toroidal coordinates from grid cell population activity and employs path-lifting to reconstruct trajectories in physical space. The reconstructed paths differ from the original by an affine transformation. We validated the method on both continuous attractor network simulations and experimental recordings of grid cells, demonstrating that local trajectories can be reliably reconstructed from a single grid cell module without external position information or training data. These results suggest that co-modular grid cells contain sufficient information for path integration and suggest a potential computational mechanism for spatial navigation.

Phenotypic heterogeneity is a pervasive feature of biological populations, yet its impact on population growth is often interpreted through changes in mean individual fitness alone. In this paper, we investigate how nonheritable variability in division rates influences the asymptotic growth of a population. Using a class of linear models with phenotypic structure, we show that variability modifies the dominant eigenvalue of the system in a nonlinear manner, leading to an intrinsic tradeoff: while variability reduces population growth under favourable conditions, it mitigates population decline under stress. These results provide a simple mechanistic framework for understanding how heterogeneity influences population-level dynamics. In particular, they suggest that stress-dependent amplification of mutational effects may arise from changes in phenotypic variability, rather than from changes in mean fitness alone. We illustrate this mechanism using mutation accumulation data in Chlamydomonas reinhardtii, where the observed patterns of relative fitness under increasing stress are consistent with increased variability within genotypes. More broadly, our analysis highlights the importance of variability as a determinant of population growth, and shows that some effects commonly attributed to changes in mean fitness may instead reflect the nonlinear consequences of phenotypic heterogeneity.

In mathematical phylogenetics, evolutionary relationships are often represented by trees and networks. The latter are typically used whenever the relationships cannot be adequately described by a tree, which happens when reticulate evolutionary events happen, such as horizontal gene transfer or hybridization. But as such events are known to be relatively rare for most species, evolution is sometimes thought of as a process that can be represented by a tree with some additional edges, i.e., with a network that is still "somewhat tree-like". In this context, different versions of tree-based networks have played a major role in recent phylogenetic literature. Yet, surprisingly little is known about their combinatorial and graph-theoretic properties. In our manuscript, we answer a recently published question concerning the colorability of a specific class of tree-based networks. In particular, we will investigate an even more general class of graphs and show their 3-colorability. This nicely links recent phylogenetic concepts with classical graph theory. Moreover, the ideas we use to answer the colorability question are new and might potentially be generalizable to other coloring problems in graph theory.

Amplifying weak molecular signals is essential in both natural and engineered biochemical systems. While most amplification schemes operate out of equilibrium, relying on kinetic barriers and fuel-driven cascades, it is also possible to amplify at thermodynamic equilibrium by shifting the energy landscape upon addition of an analyte. Equilibrium amplification is appealing because, in principle, it can remain indefinitely in the untriggered state. In this work, we establish fundamental structural and thermodynamic limits on equilibrium-based amplification. We first prove that dimerization networks--systems restricted to complexes of at most two monomers--are inherently incapable of equilibrium amplification. This no-go theorem explains the absence of amplification in prior undercomplementary "strand commutation" designs. We then show that allowing trimeric complexes breaks this barrier. We propose an isometric trimer-based amplifier whose output preserves the size of the input, enabling modular composition, and validate it experimentally, achieving an amplification factor close to the expected $2\times$. Finally, we derive universal thermodynamic bounds applicable to any equilibrium network regardless of complex size: the maximum amplification factor scales linearly with the free energy of interaction between the analyte and the amplifier components. For nucleic acid systems, this implies that the analyte length must grow linearly with the desired amplification factor, and that composing modular amplifiers yields diminishing returns for a fixed analyte. Together, these results delineate the structural and energetic boundaries of equilibrium amplification and rigorously justify the necessity of out-of-equilibrium approaches for achieving high gain.

The sensory cortices of the brain perform perceptual inference efficiently through their complex networks of neurons. One of the theoretical accounts of this process is the free-energy principle (FEP), which postulates that the brain performs variational Bayesian inference. Pioneering studies have shown that FEP can correspond to the predictive coding (PC) hypothesis under the Gaussian assumption and Laplace approximation. However, PC-based implementations of FEP within such a limited Gaussian regime have failed to capture several properties of biological neural networks, such as nonlinearity and heterogeneity of input--output properties within a network, and the biological implausibility of negative firing rates. This study shows that, when a broader class of probability distributions, namely the exponential family of distributions (EFD), is assumed for the variational posterior and prior, these missing characteristics are exhibited within the network, maintaining the FEP--PC correspondence up to the second cumulant of the posterior. We also show that the proposed model can be trained by biologically plausible local plasticity rules. Our results enrich the explanatory power of FEP regarding neural dynamics involved in perception as variational inference.

Markerless motion tracking has advanced rapidly in the past 10 years and currently offers powerful opportunities for behavioural, clinical, and biomechanical research. While several specialised toolkits provide high performance for specific tasks, using existing tools still requires substantial technical expertise. There remains a gap in accessible, integrated solutions that deliver sufficient tracking for non-experts across diverse settings. TrackStudio was developed to address this gap by combining established open-source tools into a single, modular, GUI-based pipeline that works out of the box. It provides video recording preprocessing, recording synchronisation, automatic 2D and 3D pose estimation, and visualisation without requiring any programming skills. We supply a user guide with practical advice for video acquisition, camera calibration, video synchronisation, and experimental setup, alongside documentation of common pitfalls and how to avoid them. To validate the toolkit, we tested its performance across three environments using either low-cost webcams or high-resolution cameras, including challenging conditions for body position, lighting, space, and obstructions. Across 76 participants, average inter-frame correlations exceeded 0.98 and average triangulation errors remained low (<13.6mm for hand tracking), demonstrating stable and consistent tracking. We further show that the same pipeline can be extended beyond hand tracking to other body and face regions. TrackStudio provides a practical, accessible route into markerless tracking for researchers or laypeople who need reliable performance without specialist expertise.

Vision foundation models trained with self-supervised objectives achieve strong performance across diverse tasks and exhibit emergent object segmentation properties. However, their alignment with human object perception remains poorly understood. Here, we introduce a behavioral benchmark in which participants make same/different object judgments for dot pairs on naturalistic scenes, scaling up a classical psychophysics paradigm to over 1000 trials. We test a diverse set of vision models using a simple readout from their representations to predict subjects' reaction times. We observe a steady improvement across model generations, with both architecture and training objective contributing to alignment, and transformer-based models trained with the DINO self-supervised objective showing the strongest performance. To investigate the source of this improvement, we propose a novel metric to quantify the object-centric component of representations by measuring patch similarity within and between objects. Across models, stronger object-centric structure predicts human segmentation behavior more accurately. We further show that matching the Gram matrix of supervised transformer models, capturing similarity structure across image patches, with that of a self-supervised model through distillation improves their alignment with human behavior, converging with the prior finding that Gram anchoring improves DINOv3's feature quality. Together, these results demonstrate that self-supervised vision models capture object structure in a behaviorally human-like manner, and that Gram matrix structure plays a role in driving perceptual alignment.

Collective oscillations in neuronal systems often arise from interactions between excitatory and inhibitory populations rather than from recurrent coupling within a single ensemble. Motivated by the coexistence of strongly and partially synchronized regimes in such systems, we study the Kuramoto Sakaguchi model on a bipartite network. Despite its minimal structure, the model exhibits rich collective dynamics, including both continuous and discontinuous transitions from full synchrony to partial synchrony (PS). In the PS regime, global oscillations fail to entrain one of the two populations, whose oscillators display quasiperiodic dynamics with an average frequency that can significantly deviate from that of the global field, as observed in neuronal networks. We show that this PS state constitutes an example of self-organized quasiperiodicity, arising here in the canonical Kuramoto Sakaguchi model despite its purely linear global coupling.