Strongly coupled, recurrent, balanced network models have been successful in describing and predicting many phenomena observed in cortical neural recordings. However, most balanced network models use current-based synapse models in place of more realistic, conductance-based models. Conductance-based synapse models predict unrealistically small membrane potential variability. On the other hand, introducing realistic levels of spike time correlations to models with current-based synapses predicts unrealistically large membrane potential variability. We use computer simulations to show that these two effects can cancel: Recurrent network models with conductance-based synapses and spike time correlations produce more realistic, moderate levels of membrane potential variability. Consistent with recent work on feedforward networks, our results show that including more realistic modeling assumptions produces more realistic dynamics, but only if when two modeling assumptions are included together.

Microbiome functions are encoded within the genes of the community-wide metagenome. A natural question is whether properties of a microbial community can be predicted just from knowing the raw DNA sequences of its members. In this work, we employ set-aggregated genome embeddings (SAGE) to predict community-level abundance profiles, exploiting the few-shot learning capabilities of genomic language models (GLMs). We benchmark this approach to show improved generalization on novel genomes compared to classical bioinformatics approaches. Model ablation shows that community-level latent representations directly result in improved performance. Lastly, we demonstrate the benefits of intermediate transformations between latent representations and demonstrate the differences between GLM embedding choices.

Periodic signals are critical for representing physical and perceptual phenomena. Scalar, real angular measures, e.g., radians and degrees, result in difficulty processing and distinguishing nearby angles, especially when their absolute difference exceeds pi. We can avoid this problem by using real-valued, periodic embeddings in high-dimensional space. These representations also allow us to control the nature of their dot product similarities, allowing us to construct a variety of different kernel shapes. In this work, we aim of highlight how these representations can be constructed and focus on the formalization of Dirichlet and periodic Gaussian kernels using the neurally-plausible representation scheme of Spatial Semantic Pointers.

Genome engineering has achieved remarkable sequence-level precision, yet predicting the transcriptomic state that a cell will occupy after perturbation remains an open problem. Single-cell CRISPR screens measure how far cells move from their unperturbed state, but this effect magnitude ignores a fundamental question: do the cells move together? Two perturbations with identical magnitude can produce qualitatively different outcomes if one drives cells coherently along a shared trajectory while the other scatters them across expression space. We introduce a geometric stability metric, Shesha, that quantifies the directional coherence of single-cell perturbation responses as the mean cosine similarity between individual cell shift vectors and the mean perturbation direction. Across five CRISPR datasets (2,200+ perturbations spanning CRISPRa, CRISPRi, and pooled screens), stability correlates strongly with effect magnitude (Spearman $ρ=0.75-0.97$), with a calibrated cross-dataset correlation of 0.97. Crucially, discordant cases where the two metrics decouple expose regulatory architecture: pleiotropic master regulators such as CEBPA and GATA1 pay a "geometric tax," producing large but incoherent shifts, while lineage-specific factors such as KLF1 produce tightly coordinated responses. After controlling for magnitude, geometric instability is independently associated with elevated chaperone activation (HSPA5/BiP; $ρ_{partial}=-0.34$ and $-0.21$ across datasets), and the high-stability/high-stress quadrant is systematically depleted. The magnitude-stability relationship persists in scGPT foundation model embeddings, confirming it is a property of biological state space rather than linear projection. Perturbation stability provides a complementary axis for hit prioritization in screens, phenotypic quality control in cell manufacturing, and evaluation of in silico perturbation predictions.

Actin flow in the cortical cytoskeleton underneath the cell membrane generates mechanical stresses that shape the cell surface. We study this mechanism using a hydrodynamic model of a compressible active gel polymerizing at the membrane and undergoing turnover. We determine how actin flow, density relaxation and friction of actin with the membrane generate stress on a corrugated membrane at the linear order in deformation. Analytical solutions in limiting regimes, combined with finite element methods in the general case, provide a map of normal and tangential stresses as functions of compressibility, interfacial friction and actin turnover, and determine the conditions under which actin polymerization can render the membrane linearly unstable. The non-linear regime is also briefly discussed.

We introduce DeepFHT, a survival-analysis framework that couples deep neural networks with first hitting time (FHT) distributions from stochastic process theory. Time to event is represented as the first passage of a latent diffusion process to an absorbing boundary. A neural network maps input variables to physically meaningful parameters including initial condition, drift, and diffusion, within a chosen FHT process such as Brownian motion, both with drift and driftless. This yields closed- form survival and hazard functions and captures time-varying risk without assuming proportional- hazards. We compare DeepFHT with Cox regression using synthetic and real-world datasets. The method achieves predictive accuracy on par with the state-of-the-art approach, while maintaining a physics- based interpretable parameterization that elucidates the relation between input features and risk. This combination of stochastic process theory and deep learning provides a principled avenue for modeling survival phenomena in complex systems

Phase separation in biomolecular mixtures can result from multiple physical interactions, which may act either complementarily or antagonistically. In the case of protein-nucleic acid mixtures, charge plays a key role but can have contrasting effects on phase behavior. Attractive electrostatic interactions between oppositely charged macromolecules are screened by added salt, reducing the driving force for coacervation. By contrast, base pairing interactions between nucleic acids are diminished by charge repulsion and thus enhanced by added salt, promoting associative phase separation. To explore this interplay, we combine experiment and theory to map the complex phase behavior of a model solution of poly-L-lysine (PLL) and self-complementary DNA nanostars (NS) as a function of temperature, ionic strength, and macromolecular composition. Despite having opposite salt dependences, we find that electrostatics and base pairing cooperate to stabilize NS-PLL coacervation at high ionic strengths and temperatures, leading to two- or three-phase coexistence under various conditions. We further observe a variety of kinetic pathways to phase separation at different salt concentrations, resulting in the formation of nonequilibrium aggregates or droplets whose compositions evolve on long timescales. Finally, we show that the cooperativity between electrostatics and base pairing can be used to create immiscible coacervates that partition various NS species at intermediate salt concentrations. Our results illustrate how the interplay between distinct interaction modes can greatly increase the complexity of the phase behavior relative to systems with a single type of interaction.

Neurons rely on two interdependent mechanisms, homeostasis and neuromodulation, to maintain robust and adaptable functionality. Calcium homeostasis stabilizes neuronal activity by adjusting ionic conductances, whereas neuromodulation dynamically modifies ionic properties in response to external signals carried by neuromodulators. Combining these mechanisms in conductance-based models often produces unreliable outcomes, particularly when sharp neuromodulation interferes with calcium-homeostatic tuning. This study explores how a biologically inspired neuromodulation controller can harmonize with calcium homeostasis to ensure reliable neuronal function. Using computational models of stomatogastric ganglion and dopaminergic neurons, we demonstrate that controlled neuromodulation preserves neuronal firing patterns while calcium homeostasis simultaneously maintains target intracellular calcium levels. Unlike sharp neuromodulation, the neuromodulation controller integrates activity-dependent feedback through mechanisms mimicking G-protein-coupled receptor cascades. The interaction between these controllers critically depends on the existence of an intersection in conductance space, representing a balance between target calcium levels and neuromodulated firing patterns. Maximizing neuronal degeneracy enhances the likelihood of such intersections, enabling robust modulation and compensation for channel blockades. We further show that this controller pairing extends to network-level activity, reliably modulating the rhythmic activity of central pattern generators. This study highlights the complementary roles of calcium homeostasis and neuromodulation, proposing a unified control framework for maintaining robust and adaptive neural activity under physiological and pathological conditions.

Advanced microscopy techniques are essential in biomedical research for visualising and tracking biomolecules within living cells and their compartments. Conventional fluorescence microscopy methods, however, often struggle with accurately measuring the absolute concentrations of fluorescent probes in living cells. To overcome these limitations, we introduce an open-source analysis tool, smICA (Single-Molecule Image to Concentration Analyser). The smICA method offers quantitative mapping of absolute fluorophore concentrations, lifetime-resolved filtering methods of the signal, intensity-based cell segmentation, and requires only a few photons per pixel. Our approach also reduces the time required to determine the mean concentration per cell compared to the standard FCS measurement performed in multiple posts. To highlight the robustness of the method, we validated it against standard fluorescence correlation spectroscopy (FCS) measurements by performing in vitro (polymers in aqueous solution) and in vivo (polymers and EGFP in living cells) experiments. Finally, we present exemplary studies on the time evolution of fluorescently labelled mRNA concentration in living cells. The presented methodology, along with the software, is a promising tool for quantitative single-cell studies, including, but not limited to, protein expression, biomolecule degradation (such as proteins and mRNA), and monitoring enzymatic reactions.

Neuronal systems maintain stable functions despite large variability in their physiological components. Ion channel expression, in particular, is highly variable in neurons exhibiting similar electrophysiological phenotypes, which poses questions regarding how specific ion channel subsets reliably shape neuron intrinsic properties. Here, we use detailed conductance-based modeling to explore the origin of stable neuronal function from variable channel composition. Using dimensionality reduction, we uncover two principal dimensions in the channel conductance space that capture most of the variance of the observed variability. Those two dimensions correspond to two physiologically relevant sources of variability that can be explained by feedback mechanisms underlying regulation of neuronal activity, providing quantitative insights into how channel composition links to neuronal electrophysiological activity. These insights allowed us to understand and design a model-independent, reliable neuromodulation rule for variable neuronal populations.

Machine-learning predictors of biochemical activity often exhibit large random-split-to-leave-one-target-out generalisation gaps that have been documented but not decomposed. We frame this as an evaluation-science question and use targeted protein degradation as the empirical test bed. PROTACs (proteolysis-targeting chimeras) are heterobifunctional small molecules that induce targeted protein degradation, with more than forty candidates currently in clinical trials; published predictors report AUROC of 0.85 to 0.91 under random-split cross-validation, while the leave-one-target-out (LOTO) protocol of Ribes et al. reduces performance to approximately 0.67. Random splits reward within-target interpolation, whereas LOTO measures the novel-target prediction that de-novo design depends on. We decompose this gap and identify inter-laboratory measurement variance as the dominant component, anchored by a within-target cross-laboratory cascade bounding the inter-laboratory contribution at 0.124 AUROC, well above the 0.05 contribution from binarisation-threshold choice. Across eight published architectures and ESM-2 protein language models up to 3B parameters, LOTO AUROC plateaus near 0.67, with a comparable plateau under SMILES-level deduplication; a 21-dimensional 2000-trial hyperparameter optimisation cannot break this ceiling, and the rank-1 single-seed configuration regresses by 0.161 AUROC under multi-seed validation, matching a closed-form selection-bias prediction (Bailey and Lopez de Prado, 2014). Few-shot k=5 stratified per-target retraining combined with ADMET features lifts 65-target LOTO AUROC from 0.668 to 0.7050, and post-hoc Platt scaling recovers raw output to within the 0.05 well-calibrated threshold. We release PROTAC-Bench (10,748 measurements, 173 targets, 65 LOTO folds), the variance-decomposition framework, the per-target calibration protocol, and the evaluation code.

The spatial and functional organization of the primate visual cortex is a fundamental problem in neuroscience. While recent computational frameworks like the Topographic Deep Artificial Neural Network (TDANN) have successfully modeled spatial organization in the ventral stream, the computational origins of the dorsal stream's distinct topographies, such as direction-selective maps in the middle temporal (MT) area, remain largely unresolved. In this work, we present a spatiotemporal TDANN to investigate whether MT topography is governed by the same universal principles. By training a 3D ResNet on naturalistic videos via a Momentum Contrast (MoCo) self-supervised paradigm alongside a biologically inspired spatial loss, we demonstrate the spontaneous emergence of brain-like direction maps and topological pinwheel structures. Crucially, we reveal that MT tuning properties, characterized by strong direction selectivity paired with a residual axial component, arise from a strict optimization trade-off between task-driven discriminative pressure and spatial regularization. The model's representations quantitatively match in vivo macaque MT physiological baselines, including direction selectivity index, circular variance, and pinwheel density. These findings unify the computational origins of the ventral and dorsal streams, establishing a general mechanism for cortical self-organization.

The Inositol 1,4,5-trisphosphate receptor channel (IP 3 R) is an important calcium channel involved in calcium-induced calcium release, playing a prominent role in intracellular calcium signaling. However, accurately characterizing its gating behavior remains a challenge, particularly due to the temporal resolution of patch clamp techniques that is not large enough to detect all short-lived events. This limitation can significantly bias the inference of kinetic models describing the receptor activity. To address this issue, we focused on the quantitative analysis of IP 3 R gating behavior using patch clamp data, with particular attention to missed events. We modeled IP 3 R channel gating using Hierarchical Markov chains and used a Bayesian approach that integrates missed event correction directly into the likelihood function, enabling more accurate parameter inference and model evaluation. We show that accounting for missed events deeply clarifies the multi-modal model that emerges from model selection. In this new model, the Park and Drive modes both consist of the same 3-state Markov model, with mode-dependent kinetic parameters: the Drive mode stabilizes the closed state directly connected to the open one, whereas the Park mode stabilizes the other closed state, that is not connected to the open one. Intermediate Ca 2+ concentrations are found to strongly depress the Drive to Park transition rate, so that the IP 3 R channel undergoes frequent transitions to the Park mode only for __ 50 nM or micromolar Ca 2+ concentrations. Overall, our approach provides a refined perspective on IP 3 R channel modeling and highlights the critical importance of accounting for missed events upon model selection based on single-channel recordings.

NORI performs probabilistic inference to resolve ambiguous mappings between experimental observations and biological entities orders of magnitude faster than state-of-the-art methods. This makes large-scale analysis and extensive hyperparameter optimization possible, and supports a broader range of bioinformatics applications, including protein inference, taxonomic and functional analysis in omics-fields.

The increasing adoption of data-driven decision-making in public health has established epidemic forecasting as a critical area of research. Recent advances in multivariate forecasting models better capture complex temporal dependencies than conventional univariate approaches, which model individual series independently. Despite this potential, the development of robust epidemic forecasting methods is constrained by the lack of high-quality benchmarks comprising diverse multivariate datasets across infectious diseases and geographical regions. To address this gap, we present EpiCastBench, a large-scale benchmarking framework featuring 40 curated (correlated) multivariate epidemic datasets. These publicly available datasets span a wide range of infectious diseases and exhibit diverse characteristics in terms of temporal granularity, series length, and sparsity. We analyze these datasets to identify their global features and structural patterns. To ensure reproducibility and fair comparison, we establish standardized evaluation settings, including a unified forecasting horizon, consistent preprocessing pipelines, diverse performance metrics, and statistical significance testing. By leveraging this framework, we conduct a comprehensive evaluation of 15 multivariate forecasting models spanning statistical baselines to state-of-the-art deep learning and foundation models. All datasets and code are publicly available on Kaggle (https://www.kaggle.com/datasets/aimltsf/epicastbench) and GitHub (https://github.com/aimltsf/EpiCastBench).

Topological data analysis (TDA) has established itself as a useful tool for capturing multiscale structures in complex networks, such as connected components, cycles, and cavities. Although Vietoris-Rips (VR) filtering is widely used in network analysis, it tends to be computationally expensive, especially for large networks. This work explores vertex function-based (VFB) filtering based on network measures, applying persistent homology to identify relevant topological structures in cancer-associated protein networks, and compares its effectiveness with the VR approach. The results show that VFB reproduces the second-order structures (Betti-2) identified by VR, recovering previously reported essential genes. In addition, VFB detected new driver genes, confirmed in databases such as IntOGen and NCG, and allowed analysis of third-order structures (Betti-3) that was not feasible with VR. Thus, VFB represents a scalable alternative to VR, preserving biological interpretability and complementing classical network metrics.

Bifunctional enzymes, which catalyze both the forward and reverse steps of a substrate modification reaction, arise naturally in bacterial two-component signaling systems and metabolic regulation. Beyond their well-known role in conferring absolute concentration robustness (ACR) on substrate species, bifunctional enzymes profoundly shape the dynamical landscape of the networks in which they appear. We study a class of dual-site futile cycles in which the reverse modification steps are carried out by bifunctional enzyme-substrate compounds, and provide a complete mathematical analysis of all four such networks, characterizing the existence, number, and stability of steady states, as well as the bifurcation structure as total substrate is varied. All four networks admit boundary steady states, in contrast to the non-bifunctional case. The networks differ in the number and stability of boundary steady states, in the maximum number of positive steady states (ranging from two to four), and in whether bistability is present. In two networks, a transcritical bifurcation connects the boundary and positive steady state branches; in one case this is a backward bifurcation, producing hysteresis. Perhaps the most striking phenomenon occurs in one of the four networks, which simultaneously exhibits bistability and ACR in the final modification state, where the system can settle into either of two stable steady states with different intermediate concentrations yet identical final product concentration.

We study the application of recent Edit Flows for inference-time reward control for DNA sequence generation. Unlike most reward-guided DNA generation frameworks, which operate on fixed-length sequence spaces, Edit Flows have a potential to generate variable-length DNA through biologically plausible insertion, deletion, and substitution operations. In particular, we propose Local Perturbation Discrete Programming (LPDP), a training-free, intermediate-state and action-aware local re-solving operator for variable-length DNA edit-action generators at inference time. More specifically, at each guided rollout step, LPDP scores one-step root edits, retains a near-best root band, and re-ranks each retained root by solving a bounded local discrete program around its child sequence. This local program uses the typed geometry of edit actions to focus on coherent substitution, insertion, or deletion subgraphs, and aggregates local continuations with either a hard Max backup or a soft log-sum-exponential (LSE) backup. We instantiate LPDP in two regimes: front-loaded reward tilting for enhancer optimization, where early edits are critical for establishing global regulatory sequence structure, and back-loaded reward tilting for exon-intron-exon inpainting, where late edits fine-tune splice-boundary contexts.

Autonomous science promises to augment scientific discovery, particularly in complex fields like biomedicine. However, this requires AI systems that can consistently generate novel and diverse solutions to open-ended problems. We evaluate LLMs on the task of open-ended solution generation and quantify their tendency to mode collapse into low-diversity generations. To mitigate this mode collapse, we introduce analogical reasoning (AR) as a new approach to solution generation. AR generates analogies to cross-domain problems based on shared relational structure, then uses those analogies to search for novel solutions. Compared to baselines, AR discovers significantly more diverse generations (improving solution diversity metrics by 90-173%), generates novel solutions over 50% of the time (compared to as little as 1.6% for baselines), and produces high-quality analogies. To validate the real-world feasibility of AR, we implement AR-generated solutions across four biomedical problems, yielding consistent quantitative gains. AR-generated approaches achieve a nearly 13-fold improvement on distributional metrics for perturbation effect prediction, outperform all baselines on AUPRC when predicting cell-cell communication, infer brain region interactions with a high Spearman correlation ($ρ$=0.729) to published methods, and establish state-of-the-art performance on 2 datasets for oligonucleotide property prediction. The novel and diverse solutions produced by AR can be used to augment the search space of existing solution generation methods.

Predictive models in biomedicine depend on structured assay data locked in the text, tables, and supplements of primary publications. This bottleneck is especially acute in targeted protein degradation (TPD), where each assay record must combine compound identity, degradation target, recruiter, assay context, and endpoint values reported across sections, tables, and supplementary files. Inconsistent compound identifiers and incomplete or implicit assay context further demand domain-specific logic that generic LLM pipelines do not provide. Existing molecular glue and PROTAC databases are manually curated and often lack the experimental context required for downstream modeling. We formulate TPD database extraction as a domain-specific curation task and present an expert-in-the-loop LLM workflow, evaluated through a triangular comparison among LLM predictions, standardized baseline records, and expert-annotated ground truth. A lightweight cross-validated prompt-refinement module adapts extraction instructions from scarce expert annotations. With only seven annotated molecular glue publications, the workflow achieved record-level $F_1 = 0.98$ and transferred to PROTACs by terminology substitution alone, maintaining record-level $F_1 > 0.93$. Applied at scale, it expanded molecular glue and PROTAC databases by 81% and 92% records, respectively, with 92% and 82.5% of newly recovered records validated as correct upon expert review. The workflow also recovered kinetic and assay-context information essential for cross-study potency comparison and condition-aware degradation modeling. We release the workflow, prompts, evaluation code, and extracted datasets as resources for TPD data curation and AI-assisted scientific curation more broadly.

Accurately modeling and designing protein complex structures is a central problem in computational structural biology, with broad implications for understanding cellular function and developing therapeutics. This thesis investigates two fundamental aspects of this problem using deep learning: domain-specific architectures that capture the hierarchical nature of protein structures, and search algorithms that efficiently navigate the vast sequence spaces of protein complexes to identify interacting homologs for improving complex structure prediction and to design protein sequences.

One of the great phytogeography zones of semi-arid lands in the world is the Kurdistan region of Iraq which hosts many important fruit species due to its geographical location and ecology. Mountain Hawthorn (Crataegus spp.) is a vital wild edible deciduous fruit tree of the genus Crataegus for the region, which is highly beneficial for ornamental, economical, industrial and medicinal uses. In the present study, morphological, phytochemical and molecular marker systems were applied on sixty-one Hawthorn accessions from different locations in the Iraqi Kurdistan region during April 2022 to September 2023. Phenotypic markers have proven to be extremely useful in studies of genetic diversity in Hawthorn genotypes, the results of the present morphological study showed that there are seven taxa (five species, two hybrids) were observed including, Crataegus azarolus, Crataegus meyrei, Crataegus monogyna, Crataegus orientalists, Crataegus pentagyna, Crataegus azarolus x Crataegus meyrei and Crataegus azarolus x Crataegus pentagyna. There was significant variation among different ecotypes in terms of plant type, reproductive stage, and fruit morphology and production uses. Fruit Physio-morphological data revealed a high level of significant variability (P 0.01) among accessions based on the analysis of variance. The most important characteristics for explaining fruit morphological variability `were 11 varbales including fruit weight (FW), fruit length (FL), fruit width (FW), seed length (SL), seed width (SW), number of seeds per fruits (NSF), volume solution (VS), fruit fresh weight (WOF), seed weight (WS), Potentional of hydrogen (pH) and mositure content (MC). They all are significantly different for all the traits measured for the studied accessions.

Identifying operons is a fundamental step in understanding prokaryotic gene regulation, as classifying genes into operons supports the reconstruction of regulatory networks, functional annotation of unannotated genes, and drug candidate development. Experimental approaches such as RT-PCR and RNA-seq provide precise evidence of operon structure, but are laborious and largely limited to well-studied model organisms, making scalable computational methods essential for genome-wide operon identification. Prior computational approaches have employed traditional classifiers such as logistic regression and decision trees, motivating our use of these as physicochemical baselines. The DGEB benchmark evaluates operonic pair classification by embedding each sequence independently with a pre-trained protein language model and computing pairwise cosine similarity. In contrast, our Siamese MLP learns a classifier over the fused embedding space, which is theoretically better motivated for binary classification, as cosine similarity can yield meaningless scores depending on the regularization of the embedding model. While protein language model embeddings substantially outperform physicochemical features in ROC-AUC, a learned Siamese MLP head does not significantly improve over unsupervised cosine similarity in Average Precision, suggesting that the geometry of the embedding space already captures the functional relationships needed for this task. Nonetheless, our Siamese MLP achieves a ROC-AUC of 0.71, competitive with state-of-the-art models on the DGEB leaderboard. These findings indicate that protein language model embeddings are a viable, scalable foundation for operonic pair classification across diverse microbial genomes, with implications for automated genome annotation, regulatory network reconstruction, and characterization of organisms lacking experimental operon annotations.

Learning in artificial neural networks usually relies on continuous, externally driven weight updates, in which parameters are modified at every step in response to incoming data, error signals or reward feedback. In this setting, routine and informative inputs contribute similarly to parameter adjustment. We introduce a learning approach in which parameter updates are governed by internally generated events arising from the network own representational dynamics. During ongoing activity, synaptic interactions are accumulated as latent traces encoding recent coactivation patterns, without immediately modifying the underlying parameters. In parallel, an internal predictive process estimates the evolving latent state, while a scalar measure of discrepancy between predicted and observed states is continuously computed. When discrepancy exceeds an adaptive threshold derived from recent error statistics, a learning event is triggered, inducing a retrospective update selectively integrating past activity into the current configuration. We performed simulations using a minimal neural network exposed to structured sequential inputs with transient perturbations. We found that learning occurs through sparse, temporally localized events associated with increases in prediction error, leading to stepwise changes in synaptic efficacy and discrete transitions in latent state organization. By selectively reorganizing parameters in response to internally detected discrepancies, our episodic updating may reduce unnecessary parameter drift while preserving informative patterns. Potential applications include systems requiring selective adaptation to rare or informative inputs such as physiological, industrial or environmental monitoring, edge computing under limited energy budgets, autonomous systems operating in dynamic conditions and sequential computational data processing.

Protein language models such as ESM-2 learn rich residue representations that achieve strong performance on protein function prediction, but their features remain difficult to interpret as structural $\&$ evolutionary signals are encoded in dense latent spaces. We propose a plug-$\&$-play framework that projects ESM-2 representations onto protein contact graphs $\&$ applies $\textbf{SoftBlobGIN}$, a lightweight Graph Isomorphism Network with differentiable Gumbel-softmax substructure pooling, to perform structure-aware message passing $\&$ learn coarse functional substructures for downstream prediction tasks. Across enzyme classification, SoftBlobGIN achieves 92.8\% accuracy $\&$ 0.898 macro-F1. Unlike post hoc analysis of protein language models alone, our method produces directly auditable structural explanations: GNNExplainer recovers biologically meaningful active-site residues, spatially localized functional clusters, $\&$ catalytic contact patterns. On binding-site detection, SoftBlobGIN improves residue AUROC from $0.885$ using an ESM-2 linear probe to $0.983$, indicating that these structural explanations are not recoverable from language-model features alone. Learned blob partitions provide an additional layer of interpretability by automatically grouping residues into functional substructures, with blobs containing annotated active-site residues showing $1.85\times$ higher importance than other blobs ($ρ{=}0.339$, $p{=}0.009$), without any active-site supervision. Our framework requires no retraining of the language model, adds only $\sim$1.1M parameters, $\&$ generalises across ProteinShake tasks, achieving $F_{\max}$ of $0.733$ on Gene Ontology prediction $\&$ AUROC of $0.969$ on binding-site detection. We position this as an interpretable structural companion to protein language models that makes their predictions more transparent $\&$ auditable.

Importance: The 2026 multisociety dyslipidemia guideline recommended the PREVENT equations in place of the PCE equations, introduced 30-year risk assessment as a new treatment pathway, and lowered risk-based treatment thresholds. The net population impact of these concurrent changes on statin recommendations is unknown. Objective: To estimate changes in statin recommendations under 2026 PREVENT-based dyslipidemia guidelines compared with 2018 PCE-based guidelines. Design and Participants: Cross-sectional analysis of pooled data from NHANES, spanning 2011-2023 and comprising 24,199 participants aged 30-79 years. Main Outcomes and Measures: Number and proportion of US adults receiving or recommended for statin therapy. Results: At the class 1 threshold, the number of US adults receiving or recommended for statin therapy decreased by an estimated 3.0 million (95% CI, 2.3 million to 3.6 million), with larger reductions among Black adults (-4.2 percentage points [pp]), men (-4.0pp), and adults aged 50-69 years (-5.6pp). At the class 2 threshold--which additionally recommends statins for adults aged 30-59 years based on 30-year risk--the number of adults recommended increased by an estimated 20.8 million (95% CI, 19.6 million to 22.0 million), or +11.6pp. The increase was largest among adults aged 50-59 years (+19.7pp) and 40-49 years (+14.8pp). Conclusions: The net population impact of the 2026 dyslipidemia guidelines depends critically on which recommendation class is applied. At the class 1 threshold, statin recommendations decreased modestly; at the class 2 threshold, inclusion of 30-year risk assessment substantially expanded recommendations, particularly among younger adults. These divergent effects underscore the importance of the 30-year risk criterion as a major driver of new eligibility and the need for outcomes and equity monitoring during guideline implementation.

We present Clin-JEPA, a multi-phase co-training framework for joint-embedding predictive (JEPA) pretraining on EHR patient trajectories. JEPA architectures have enabled latent-space planning in robotics and high-quality representation learning in vision, but extending the paradigm to EHR data -- to obtain a single backbone that simultaneously forecasts patient trajectories and serves diverse downstream risk-prediction tasks without per-task fine-tuning -- remains an open challenge. Existing JEPA frameworks either discard the predictor after pretraining (I-JEPA, V-JEPA) or train it on a frozen pretrained encoder (V-JEPA 2-AC), leaving the encoder unaware of the rollout signal that the retained predictor must use at inference; co-training the encoder and predictor under a shared JEPA prediction objective would supply this grounding, but naïve co-training is unstable, with representation collapse and online/target drift causing autoregressive rollout to diverge. Clin-JEPA's five-phase pretraining curriculum -- predictor warmup, joint refinement, EMA target alignment, hard sync, and predictor finalization -- addresses each failure mode by phase, stably co-training a Qwen3-8B-based encoder and a 92M-parameter latent trajectory predictor. On MIMIC-IV ICU data, three independent evaluations support the framework: (1) latent $\ell_1$ rollout drift uniquely converges ($-$15.7%) over 48-hour horizons while baselines and ablations diverge (+3% to +4951%); (2) the encoder learns a clinically discriminative latent geometry (deteriorating-patient cohorts displace 4.83$\times$ further than stable patients in latent space, vs $\leq$2.62$\times$ for baseline encoders); (3) a single backbone outperforms strong tabular and sequence baselines on multi-task downstream evaluation. Clin-JEPA achieves mean AUROC 0.851 on ICareFM EEP and 0.883 on 8 binary risk tasks (+0.038 and +0.041 vs baseline average).

Protein-protein interactions (PPIs) are fundamental to cellular function and disease mechanisms. Current learning-based PPI predictors focus on learning powerful protein representations but neglect designing specialized classification heads. They mainly rely on generic aggregating methods like concatenation or dot products, which lack biological insight. Motivated by the biological "L3 rule", where multiple length-3 paths between a pair of proteins indicate their interaction likelihood, our study addresses this gap by designing a biologically informed PPI classifier. In this paper, we provide empirical evidence that popular PPI datasets strongly support the L3 rule. We propose an L3-path-regularized graph prompt learning method called L3-PPI, which can generate a prompt graph with virtual L3 paths based on protein representations and controls the number of paths. L3-PPI reformulates the classification of protein embedding pairs into a graph-level classification task over the generated prompt graph. This lightweight module seamlessly integrates with PPI predictors as a plug-and-play component, injecting the interaction prior of complementarity to enhance performance. Extensive experiments show that L3-PPI achieves superior performance enhancements over advanced competitors.

Long-context DNA models are limited by token-mixing cost and by how compression allocates representational budget across the genome. Existing approaches operate close to base-pair resolution, apply fixed downsampling, or learn content-dependent chunks without an explicit genomic budget, making long-context pretraining expensive and difficult to control. We introduce GeneZip, a region-aware DNA compression framework that combines H-Net-style dynamic routing with a Region-Aware Ratio (RAR) objective and bounded routing. GeneZip uses static gene-structure annotations during compression training to specify region-wise base-pairs-per-token (BPT) targets; at inference time, it compresses raw unseen DNA without annotations. GeneZip provides three main benefits. First, it is effective: GeneZip variants achieve the best validation PPL among encoder-based compressors, with GeneZip-70M operating at 137.6 BPT, and across four reproducible DNALongBench tasks--contact map prediction, eQTL prediction, enhancer-target gene prediction, and transcription-initiation signal prediction--GeneZip obtains the best average rank among compared sequence models. Second, it is redundancy-aware: a post-hoc RepeatMasker/TRF analysis shows that, without repeat supervision, GeneZip assigns higher local BPT to TE-derived interspersed repeats and tandem repeats, two major classes of repetitive DNA sequence redundancy. Third, it is efficient: by reducing the effective token-mixing length, GeneZip enables longer-context and larger-capacity pretraining, including 128K-context and 636M-parameter variants on a single A100 80GB GPU, and fine-tunes the eQTL task 50.4x faster than JanusDNA (50 vs. 2520 minutes). These results establish GeneZip as an effective, redundancy-aware, and efficient compression interface for long-context DNA modeling.

Deep generative modeling to stochastically design small molecules is an emerging technology for accelerating drug discovery and development. However, one major issue in molecular generative models is their lower frequency of drug-like compounds. To resolve this problem, we developed a novel framework for optimization of deep generative models integrated with a D-Wave quantum annealing computer, where our Neural Hash Function (NHF) presented herein is used both as the regularization and binarization schemes simultaneously, of which the latter is for transformation between continuous and discrete signals of the classical and quantum neural networks, respectively, in the error evaluation (i.e., objective) function. The compounds generated via the quantum-annealing generative models exhibited higher quality in both validity and drug-likeness than those generated via the fully-classical models, and was further indicated to exceed even the training data in terms of drug-likeness features, without any restraints and conditions to deliberately induce such an optimization. These results indicated an advantage of quantum annealing to aim at a stochastic generator integrated with our novel neural network architectures, for the extended performance of feature space sampling and extraction of characteristic features in drug design.