Biology subjects

Yao, J.

Publications and source records attributed to Yao, J..

At least 37 records · Page 2Linked to original sources

bioRxiv2026

IgGM2: An All-Atom Foundation Model for Adaptive Immune Receptor Design

Accurate immune receptor design requires modeling the coupled variation of aminoacid sequence, full-atom conformation, and target-binding geometry across antibodies, nanobodies, and T-cell receptors (TCRs). Existing methods often address only part of this problem, either by separating structure generation from sequence design, relying on fixed-backbone inverse folding, or focusing on a single receptor class. We introduce IgGM2, a unified all-atom generative framework for immune receptor structure prediction and CDR sequence-structure co-design. IgGM2 follows a structure-to-design strategy: it first learns how immune receptors are positioned around fixed target structures, and then transfers this target-conditioned structural prior to CDR design. Unlike modular design pipelines, IgGM2 jointly generates CDR residue identities and full-atom receptor structures, allowing frame-work geometry to adapt to designed CDRs without separate inverse folding or external sidechain packing. Unlike continuous residue encodings based on virtualatom geometry, IgGM2 keeps sequence prediction explicit while using atom14 placeholders only for full-atom representation. On structure prediction benchmarks, IgGM2 better captures receptor-target spatial relationships than AlphaFold3 on FoldBench and achieves strong performance on TCR-pMHC modeling. On sequence design benchmarks, IgGM2 achieves competitive amino-acid recovery and improves Rosetta-based interface preference metrics, suggesting more favorable generated binding interfaces. These results support IgGM2 as a unified all-atom framework for adaptive immune receptor structure prediction and design.

bioinformatics↗

bioRxiv2026

3D Passive Cavitation Mapping (3D-PCM) with a Large-aperture Planar Array

Urinary stone disease is a common urological condition with increasing incidence, particularly in developed countries. Laser lithotripsy (LL) has become a preferred minimally invasive treatment due to its high precision and low tissue damage. Recent studies suggest that cavitation plays a critical role in stone damage during LL, and three-dimensional passive cavitation mapping (3D-PCM) has emerged as a promising tool for detecting these events. However, clinical translation of 3D-PCM remains challenging due to limitations in imaging depth, field of view (FOV), and procedural compatibility. Here, we present a large-FOV dual-modality imaging system (3D-PCM and B-mode ultrasound) based on a large-aperture planar ultrasound array. Through array optimization and model-based reconstruction, our system achieves an expanded FOV of [~]40 x 40 mm2 at a clinically relevant imaging depth of [~]110 mm, while maintaining high spatial resolution of [~]0.6 mm laterally and [~]0.4 mm axially. In vivo experiments in a porcine model demonstrate that the reconstructed cavitation distribution correlates well with stone damage. Our technology has the potential to provide real-time treatment feedback during LL without disrupting the standard workflow.

bioengineering↗

bioRxiv2026

circPMS1-mediated LIM protein scaffolding promotes cytoskeletal remodeling and melanoma metastasis

Circular RNAs (circRNAs) have been implicated in diverse biological processes and diseases such as cancer; however, their functional roles in tumor biology remain incompletely understood. Here, we identify circPMS1 as an aberrantly upregulated, pro-metastatic circRNA in melanoma. Overexpression and silencing studies revealed that circPMS1 promotes migration and invasion of non-transformed melanocytes and melanoma cells, enhances metastasis in xenograft models, and drives spontaneous dissemination and lymph node metastases in a genetically engineered melanoma mouse model. circPMS1 pro-metastatic activity depends on its circularization and is mediated by a putative secondary structure spanning the back-splice junction and the canonical start codon of the linear PMS1 mRNA. Proteomic analysis identified LIM domain proteins, including ABLIM1, LMO7, and PDLIM7, as circPMS1-bound factors. Mechanistic studies demonstrate that circPMS1 scaffolds these LIM proteins, enhancing their association with actin filaments inducing focal adhesion and cytoskeletal remodeling. This remodeling is associated with RhoA activity, and inhibition of the RhoA pathway attenuates circPMS1-induced migration. Together, these findings establish circPMS1 as a pro-metastatic circRNA that drives melanoma progression by scaffolding LIM domain proteins and regulating cytoskeletal dynamics and cell motility.

Cancer Biology↗

bioRxiv2026

Single-cell spatial transcriptomics and snRNA-seq decoding the organizational principles of functional modules in the mouse amygdala

The amygdala is a functionally heterogeneous nuclear complex comprising multiple subnuclei that orchestrate diverse behaviors, making it essential for survival and reproduction. However, the precise neural mechanisms underlying these heterogeneous functions remain elusive, primarily due to limited knowledge of the amygdalas cellular heterogeneity, developmental origins, spatial organization, and gene expression profiles. Here, we integrate single-cell-resolution spatial transcriptomics with precise anatomical dissection and single-nucleus RNA sequencing to systematically map the subnuclear enrichment patterns of amygdalar cell types. We further demonstrate that developmental origin determines subnuclear positioning and propose a novel functional modular architecture based on cellular composition and gene expression signatures. Furthermore, we reveal that the central amygdala (CEA) exhibits developmental heterogeneity, with Isl1+ neurons in its medial subdivision (CEAm) predominantly originating from hypothalamic progenitors. Our findings establish a spatially resolved cellular and molecular framework for investigating amygdalar functions at cellular type resolution and understanding amygdala-related neuropsychiatric disorders.

neuroscience↗

bioRxiv2026

Flytrap-Inspired Mesh-Trap Bioelectronics for Full Spherical Electrophysiological Interrogation of 3D Tissues

Three-dimensional (3D) in vitro tissue models are emerging as powerful platforms for studying development, disease, and therapeutic responses, where close monitoring of electrophysiological activity is essential. However, existing probing methods remain limited in accessibility or spatiotemporal resolution for comprehensive electrophysiological mapping of suspended 3D tissues that closely mimic the native environments. Here we introduce a Venus flytrap-inspired bioelectronic mesh system that enables the full spherical enclosure of 3D tissues in a suspended configuration. The system consists of two hemispherical meshes that envelop the tissue, constructed from highly flexible, stretchable, cell-scale ribbons interconnected into a tissue-compliant network with integrated recording electrode arrays. This architecture enables intimate, conformal tissue integration and supports stable electrophysiological recordings over 300 days. The high-resolution recordings allow precise tracking of local dynamics and correlated global signaling, enabling comprehensive assessment of tissue development as well as detailed evaluation of drug responses for disease modeling. Beyond single tissues, the mesh architecture is extended to fully enclose assembloids composed of multiple tissues, enabling characterization of cross-tissue signaling relevant to advanced heterogenous tissue modeling. Furthermore, the system is translated into array-based platforms, demonstrated by a 4x4 array integrating 1024 electrodes, for high-throughput tissue sampling and cross-study analysis. The developed bioelectronic system and integration method provide a broadly applicable platform to advance electrophysiological studies across diverse tissues and organoids.

bioengineering↗

bioRxiv2026

Growth under constraints: root tip development controls trade-offs between speed and mechanical efficiency

Relationships between root tip development--particularly tip shape--are thought to play a key role in enabling plants to overcome soil mechanical resistance. However, correlations between root tip morphology and the ability to penetrate compact or hard soils have remained inconsistent. This study applies a new quantitative framework to growth kinematics data from six plant species, analysing the trade-offs between frictional energy loss, growth stability, and reduced root elongation rates. A shorter root elongation zone can reduce the fraction of the mechanical energy lost to friction, but this is done at the expense of the elongation rate. A sharper tip or increased radius can help roots maintain the elongation rate at no energetic cost, but these strategies come with the cost of growth instability (tortuous roots) and decrease in specific root length respectively. During establishment, root strategies may therefore occupy a 2-dimensional trait space in which the mechanical efficiency of growth is balanced against the explorative-exploitative trade-off. HighlightsGrowth and form of root tips explain how plants overcome mechanical resistance from the soil Trade-offs link the energy lost by friction, growth stability and elongation rate of roots Larger roots allow faster growth independently of these trade-offs New framework formalises plants strategies to acquire soil resources List of symbols O_TBL View this table: org.highwire.dtl.DTLVardef@2df502org.highwire.dtl.DTLVardef@130848eorg.highwire.dtl.DTLVardef@4ec7deorg.highwire.dtl.DTLVardef@23e5eeorg.highwire.dtl.DTLVardef@ccaae7_HPS_FORMAT_FIGEXP M_TBL C_TBL

plant biology↗

bioRxiv2026

Three-Dimensional Photoacoustic Tomography with Ultrasound Localization Priors

Three-dimensional photoacoustic tomography (3D-PAT) enables noninvasive structural and functional imaging with optical absorption contrast and ultrasonic detection depth. However, its spatial resolution is limited by acoustic diffraction, and incomplete detection geometry can substantially degrade image fidelity and quantitative accuracy. Here, we present a ULM-guided model-based reconstruction framework, termed 3D-PAULMprior that incorporates sub-diffraction vascular priors from concurrent ultrasound localization microscopy (ULM) into 3D photoacoustic reconstruction. The method uses weighted regional Laplacian regularization to integrate high-resolution vascular information into the inverse problem, thereby enhancing vascular sharpness, suppressing limited-view artifacts, and improving blood oxygen saturation estimation. We validated 3D-PAULMprior using numerical simulations, tissue-mimicking phantoms, and in vivo mouse brain imaging. Compared with conventional reconstruction, 3D- PAULMprior improved spatial resolution by over 50%, increased contrast-to-noise ratio by 261.2%, and enhanced structural similarity index by 24.6%. In vivo, 3D-PAULMprior recovered vascular structures that were poorly resolved or missing in conventional reconstructions and produced more spatially confined sO2 maps. These results establish 3D-PAULMprior as a robust multimodal reconstruction strategy for high-resolution structural and functional photoacoustic imaging.

bioengineering↗

bioRxiv2026

Protracted functional and structural reorganization of human prefrontal cortex supports lateralized category geometries

Despite being visually responsive, how prefrontal cortex represents visual categories across development and how this compares to posterior visual cortex remains unknown. Here, we show that ventrolateral prefrontal cortex (VLPFC) follows a fundamentally distinct developmental trajectory. Using functional and quantitative MRI in children and adults, we demonstrate that VLPFC exhibits reliable and structured category-selective responses in childhood that undergo prolonged reorganization into mature, hemisphere-specific representational geometries. Right VLPFC progressively organizes categories along an animate-inanimate axis, while left VLPFC consolidates a word-selective organization, a lateralization that strengthens across development. This restructuring contrasts sharply with ventral temporal cortex (VTC), which exhibits bilaterally organized, adult-like category structure in childhood. Critically, prefrontal representational structure emerges in alignment with conserved folding patterns of the precentral and inferior frontal sulci, suggesting that cortical anatomy provides a stable scaffold for the maturation of abstract representations. Quantitative MRI further reveals a non-linear trajectory of tissue development in right VLPFC, marked by adolescent reductions in macromolecular content and increased microstructural heterogeneity in adulthood, a pattern not observed in VTC. Together, these findings demonstrate that VLPFC undergoes prolonged, anatomically constrained refinement of category-related geometry characterized by hemisphere-specific taxonomies, rather than simply inheriting representational structure from sensory cortex. O_LIUnlike VTC (which shows early, bilateral, and stable category organization), VLPFC contains reliable, category-selective representations from childhood that undergo prolonged, hemisphere-specific developmental refinement rather than simple amplification C_LIO_LIRight VLPFC progressively consolidates an animate-inanimate representational axis across development, while left VLPFC establishes an early, stable word-selective organization consistent with language network maturation C_LIO_LIFunctional boundaries in VLPFC align with conserved intrasulcal folding motifs (plis de passage), demonstrating that anatomical scaffolding of category-selective responses extends beyond sensory cortex to prefrontal association cortex C_LIO_LIRight VLPFC undergoes adolescent pruning of cortical tissue microstructure diverging sharply from the continuous tissue proliferation observed in ventral temporal cortex C_LI

neuroscience↗

bioRxiv2026

3D-PAULM: Integrated Photoacoustic Tomography and Ultrasound Localization Microscopy for Multiparametric Brain and Tumor Imaging

Understanding processes such as blood-brain barrier (BBB) disruption and tumor progression can greatly benefit from simultaneous molecular, functional, and hemodynamic imaging in deep tissue, yet few existing imaging modalities can provide all three in a single system. Here, we present an integrated imaging platform that combines 3D photoacoustic tomography with ultrasound localization microscopy (3D-PAULM) to enable intrinsically co-registered, multiparametric imaging. 3D-PAULM unifies multispectral photoacoustic molecular imaging, ultrasound B-mode imaging, microbubble-enhanced power Doppler, and ultrasound localization microscopy, and concurrently measures blood oxygenation, blood perfusion, microvascular flow dynamics, and molecular probes from near-infrared dyes and photoswitchable phytochromes. We apply 3D-PAULM to quantify BBB leakage in focal ischemia and systemic inflammation, and to perform high-sensitivity molecular imaging of solid tumors alongside functional mapping of tumor hypoxia and super-resolved vascular remodeling. Together, these results establish 3D-PAULM as a versatile platform for integrated functional and molecular imaging in deep tissue.

bioengineering↗

bioRxiv2026

H2O: A Foundation Model Bridging Histopathology to Spatial Multi-Omics Profiling

Spatial omics technologies have revolutionized the molecular profiling of tissues but remain constrained by high costs and limited scalability. While hematoxylin and eosin (H&E) staining is ubiquitous, it lacks molecular specificity. Here, we present H2O (Histopathology to Omics), a generalist AI framework that bridges the modality gap between histopathology and spatial multi-omics, enabling the direct inference of spatial transcriptomics (ST) and proteomics (SP) landscapes from routine H&E images. H2O integrates Vision Transformers (ViT) with Large Language Models (LLM) via contrastive learning to align histological morphology with semantic molecular knowledge. This cross-modal approach allows the model to incorporate spatial expression profiles into histological pattern recognition, effectively decoding the molecular heterogeneity underlying tissue morphology. Trained on a pan-tissue dataset of 1.3 million paired H&E-spatial patches across 25 organs and cancer types, H2O predicts spatial omics expression from histology with high concordance to sequenced measurements and consistently outperforms state-of-the-art models across three cancer benchmarks. Notably, H2O recovers the MIF-CD74/CD44 signaling axis directly from H&E images, highlighting its capacity to infer biologically meaningful cell-cell communication without molecular profiling. Applying on three additional public cohorts covering fetal and paediatric thymus tissues, human metastatic lymph node, and breast cancer, encompassing human development, 3D spatial frameworks, and integrative multi-omics, H2O yields biologically concordant insights, demonstrating superior accuracy, robustness, and generalizability across real-world applications in diverse scenarios. H2O converts routine histopathology into a portal for spatially resolved multi-omics profiling by computationally generating transcriptomic and proteomic landscapes, thereby enhancing tissue phenotyping and enabling scalable, integrative tissue-atlas construction.

bioinformatics↗

bioRxiv2026

Photoacoustic Fingerprinting for Robust Molecular Imaging

Quantitative molecular imaging in photoacoustics is fundamentally limited by the ill-posed nature of spectral unmixing, where spectral overlap, noise, and unknown fluence introduce bias in conventional inversion-based methods. We introduce photoacoustic fingerprinting (PAF), a framework that reframes spectral unmixing as a fingerprint recognition problem. PAF interprets multispectral signals as high-dimensional fingerprints encoding both molecular composition and measurement distortions. Inspired by magnetic resonance fingerprinting, PAF uses a recurrent neural network trained on synthetic data spanning realistic mixtures, noise levels, and fluence variations to directly infer molecular concentrations from spectral shape. PAF enables accurate and robust quantification in regimes where conventional methods break down, including low signal-to-noise conditions, spectrally correlated mixtures, and unknown fluence distortions. In controlled simulations, PAF consistently outperformed non-negative least squares, with the largest gains observed for spectrally overlapping chromophores such as collagen. In phantom studies, PAF improved molecular specificity by correctly localizing collagen and recovering water contrast despite similar spectral reconstructions. In ex vivo mouse livers, PAF detected lipid accumulation associated with steatosis, and in human arteries, it identified molecular signatures consistent with thrombus and lipid-rich plaque. These results establish PAF as a generalizable framework for label-free molecular imaging and a promising step toward quantitative photoacoustic diagnostics.

biochemistry↗

bioRxiv2026

TCMCard: A High-Confidence Digital Infrastructure for Traditional Chinese Medicine Quantified by Multi-Dimensional Evidence Integration

Network pharmacology has become a widely used approach for deciphering multi-component, multi-target mechanisms of traditional Chinese medicine (TCM). Here we introduce TCMCard, a high-confidence digital infrastructure built on a Multi-Dimensional Evidence Integration (MDEI) framework. The framework integrates experimental activity data from authoritative chemical databases, literature-derived evidence, and structure-based similarity inference. Preprocessing steps include chemical structure normalization, species-specific filtering, and target quality scoring. Applied to conventional interaction datasets, this pipeline leads to the removal of over 60% of low-confidence noise. TCMCard supports network pharmacology exploration through an interactive visualization platform, and module analysis identifies functionally relevant communities that offer insights into the synergistic actions of TCM formulas. Overall, TCMCard may help move the field beyond simple data aggregation toward evidence-informed curation and quality-driven analysis. As an interactive and publicly accessible platform, it reveals an organized backbone within complex interaction networks, offering a more reliable basis for understanding multi-component synergy in TCM.

bioinformatics↗

bioRxiv2026

Sulfide:quinone oxidoreductase drives mitochondrial supersulfide metabolism to regulate bioenergetics and longevity in eukaryotes

Sulfide:quinone oxidoreductase (SQR) is a critical enzyme that maintains sulfur metabolism by oxidizing sulfide to supersulfides, currently defined as sulfur metabolites with six valence electrons and no charge that are covalently catenated with other sulfur atoms and excludes disulfides. While SQR is known to contribute to mitochondrial electron transport, its physiological impact on systemic energy metabolism and longevity remains largely undefined. In this study, we investigated the role of SQR in mitochondrial bioenergetics and aging using SQR-deficient Schizosaccharomyces pombe ({Delta}hmt2) and a mitochondria-selective SQR-deficient (Sqrdl{Delta}N/{Delta}N) mice model. Functional analysis demonstrated that{Delta} hmt2 grew normally in glucose but not in glycerol, indicating impaired mitochondrial respiration. It showed reduced membrane potential, ATP, and lifespan. Consistent with the yeast findings, Sqrdl{Delta}N/{Delta}N mice exhibited accumulated levels of hydrogen sulfide and persulfides, and demonstrated impaired mitochondrial energy metabolism. Furthermore, supersulfide donor supplementation selectively conferred lifespan extension in wild-type yeast, but not in SQR-deficient strain, and similarly improved mitochondrial function exclusively in wild-type mouse embryonic fibroblasts, with no benefit observed in SQR-mutant counterparts. Together, our findings demonstrate that mitochondrial SQR plays an essential role in sulfur respiration, critically supporting mitochondrial function and organismal longevity across eukaryotes. Graphic Abstract O_FIG O_LINKSMALLFIG WIDTH=200 HEIGHT=175 SRC="FIGDIR/small/716515v1_ufig1.gif" ALT="Figure 1"> View larger version (36K): org.highwire.dtl.DTLVardef@d834forg.highwire.dtl.DTLVardef@127dc7dorg.highwire.dtl.DTLVardef@1fccb8eorg.highwire.dtl.DTLVardef@197e910_HPS_FORMAT_FIGEXP M_FIG C_FIG HighlightsO_LIDeveloped an SQR-deficient S. pombe ({Delta}hmt2) model that exhibits sulfur metabolism, mitochondrial dysfunction, and shortened chronological lifespan C_LIO_LISulfide and supersulfide donors prolong yeast lifespan in a SQR-dependent manner C_LIO_LIMitochondrial SQR is essential for membrane potential formation and ATP production in yeast and mammals C_LI

biochemistry↗

bioRxiv2026

A feed-forward loop between niche adenosine and Gzmk⁺ CD8 T cells propagates systemic inflammaging

The causal link between the aging microenvironment and T cell aging remains elusive. Here, we demonstrate that adenosine within aging tissues actively reprograms CD8+ T cells into a pro-aging Granzyme K+ (Gzmk+) population. Mechanistically, senescent cells create an adenosine-rich niche via p16-dependent CD39 upregulation, triggering A2aR signaling to induce Gzmk+ T cell differentiation. Once released, Gzmk promotes systemic inflammaging through PAR1 and complement activation. Crucially, targeting this axis--either via genetic Gzmk ablation or pharmacological A2aR blockade--reverses multi-organ aging phenotypes and significantly extends healthy lifespan in mice. Human analysis reveals age-dependent Gzmk+ T cell accumulation in multi organs, while coffee intake (an A2aR antagonist) inversely correlates with plasma Gzmk levels. Our findings uncover how metabolic niche changes drive T cell aging and establish the adenosine-Gzmk axis as a pivotal therapeutic target for combating age-related diseases.

immunology↗

bioRxiv2026

Hypoxia-activated scleraxis a mediates epicardial progenitor differentiation into a unique cardiac perivascular cell type

The epicardium is a crucial source of progenitor cells and paracrine signals that support heart development and regeneration. However, the molecular mechanisms that guide epicardial cell fate decisions remain incompletely understood. Here, we identify the transcription factor Scleraxis a (encoded by scxa) as a key regulator of epicardial progenitor differentiation in zebrafish. Through single-cell transcriptomics, genetic lineage tracing, and cardiac injury models, we show that scxa is transiently induced in activated epicardial progenitor cells (aEPCs) during both heart regeneration and developmental coronary angiogenesis. scxa+ epicardial cells primarily give rise to a previously uncharacterized cardiac population of perivascular cells marked by col18a1a, molecularly distinct from classical pericytes and vascular smooth muscle cells. We refer to this population as epicardial-derived perivascular mesenchymal cells (Epi-PMCs). These Epi-PMCs closely associate with coronary vessels and may contribute to vascular stabilization and remodeling, potentially through the anti-angiogenic but vessel-stabilizing activity of endostatin derived from collagen XVIII. Loss of scxa increases coronary vessel density. Mechanistically, we identify hypoxia and Hif1a signaling as upstream regulators of scxa, with systemic hypoxia or Hif factor stabilization robustly inducing scxa expression in the epicardium. Together, these findings uncover a hypoxia-responsive Scxa-Col18a1a axis that drives epicardial differentiation toward a vascular-supportive fate, offering new insight into the regulation of coronary vessel development and the regenerative potential of the epicardium.

developmental biology↗

bioRxiv2026

Serpin-Driven Green Camouflage and NIR Fluorescence in Frogs

Animals have evolved multiple strategies to generate optical traits and coloration. While most amphibians rely on a three-dimensional arrangement of chromatophores in the skin, hundreds of arboreal frog species achieve leaf-like green color through a different mechanism involving reduced pigmentation, subcutaneous biological mirrors, and high concentrations of the blood-derived pigment biliverdin. Although biliverdin is rapidly excreted in most vertebrates, hylid and centrolenid frogs can retain it through the binding to a serpin-family protein (BBS). Here we show that BBSs bind biliverdin with high affinity (Kd < 10 nM), comparable to hormone-receptor interactions. This interaction alters biliverdins spectral signature in ways that resemble those of green-leaf pigments. BBSs from different species exhibit distinct biophysical properties, accounting for interspecific differences in color saturation and hue. Unlike most serpins in vivo, BBS of the glassfrog Teratohyla pulverata is naturally cleaved, yielding a highly thermostable, thermodynamically favored protein, without loss of affinity. Using custom-designed hyperspectral photoacoustic tomography (PAT), we demonstrate that BBS is widely distributed throughout the body, contributing to whole-body color and camouflage. Furthermore, we show that BBSs emit near-infrared (NIR) fluorescence (>700 nm) rendering these frogs fluorescent in a spectral region where biological tissues are largely transparent. Together, BBSs shed light on serpin evolution, protein thermostability, amphibian color diversity, and the development of NIR molecular probes.

biochemistry↗

bioRxiv2026

Single-Cell and Spatial Transcriptomics Integration Identifies Mural Cell Oxidative Stress Genes Clu and Gria2 as Key Biomarkers in Ischemic Stroke

Oxidative stress (OS) is a key factor in ischemic stroke (IS), but the characterization of OS-related genes in IS remains largely unexplored. Identifying key OS-associated genes could improve our understanding of OS in IS. We analysed single-cell RNA sequencing datasets and utilized AUCell, Ucell, singscore, ssgsea, and AddModuleScore algorithms, along with correlation analysis, identified 167 OS-related genes potentially linked to IS. Furthermore, we used seven machine learning algorithms, including least absolute shrinkage and selection operator, XGBoost, Boruta, random forest, gradient boosting machines, decision trees, and support vector machine recursive feature elimination, to identify the optimal feature genes: Clusterin (Clu) and Glutamate Ionotropic Receptor AMPA Type Subunit 2 (Gria2). Bulk RNA-sequencing showed that Clu expression was upregulated and Gria2 expression was downregulated in IS tissues. Single-cell analysis further revealed that these expression changes predominantly occurred in mural cells, emphasizing their cell-specific roles in IS. Subsequently, we validated these findings through spatial transcriptomics (ST) analysis and animal model experiments. These results suggest that Clu and Gria2 may participate in the progression of IS by altering the OS activity of mural cells. However, further research is needed to validate its clinical efficacy and ensure its application in the treatment of IS.

neuroscience↗

bioRxiv2026

MolSpecFlow: Mass-Constrained Hybrid Flow Matching for Joint Molecular-Spectral Analysis

Identifying the "dark matter" of the chemical universe requires bridging the fundamental gap between molecules and mass spectra. Existing approaches struggle to accurately map between these distinct data formats, often yielding results that are either chemically implausible or inconsistent with the physical evidence. We introduce MolSpecFlow, a unified foundation model pretrained on 100 million molecules and 42 million spectra, which leverages a hybrid flow matching framework to orchestrate optimal transport paths for spectral peaks and discrete probability flows for molecular tokens. Furthermore, to guarantee the physicochemical validity of the generated structures, we incorporate explicit rule-based constraints into the generative process, utilizing a token-level mass control mechanism that strictly enforces alignment with the precursor molecular weight. MolSpecFlow establishes a new state-of-the-art on the MassSpecGym benchmark, surpassing diffusion and retrieval baselines in de novo generation (Top-1 accuracy +35%), spectral simulation, and molecular retrieval, demonstrating the power of physics-grounded unified modeling.

bioinformatics↗