Biology subjects

Chen, S.

Publications and source records attributed to Chen, S..

At least 91 records · Page 5Linked to original sources

bioRxiv2026

The Regional Vulnerability Index (RVI) as a Neuroimaging-Based Biomarker for Autism: Associations with Likelihood, Cognition, and Longitudinal Social Outcomes

Autism spectrum disorder (ASD) is a complex neurodevelopmental condition with symptoms spanning cognitive, social and behavioral domains and leading to lifelong challenges. Autism is heritable but specific genetic and environmental factors that shape its early neurodevelopmental trajectory remain unknown. Despite the clinical variability, neuroimaging findings from Enhancing Neuro Imaging Genetics through Meta Analysis (ENIGMA)-ASD consortium identified stable and replicable patterns of cortical and subcortical differences that were consistent with those reported by an independent consortium, the Cognitive Genetics Collaborative Research Organization (COCORO) in Japan. Here we developed and evaluated a regional vulnerability index (RVI), a similarity metric that quantifies how closely a persons brain resembles the specific pattern of an individual with autism. RVI-ASD is based on combining the regional effect sizes for regional brain measurements published by the ENIGMA-ASD group with microstructural white matter integrity differences reported by COCORO. RVI-ASD showed significantly higher effect size for case-control differences relative to any individual brain measure (d=0.30 vs. d=0.01-0.21) in individuals with autism, particularly in the adolescent-to-adult sample (N=2,577; Mean age = 16.1; SD=5.0). We next calculated RVI-ASD in the baseline and follow-up (ages 10 and 12) data from normally developing participants of the ABCD study (N=4,201). We tested the longitudinal stability, heritability, genotype-by-age interaction and sensitivity of RVI-ASD to known factors and cognitive measurements. RVI-ASD were stable on the 2-year follow up (ICC=0.76-0.92); showed significant heritability (h{superscript 2}=0.55-0.83, p<10-16) but was not affected by gene-by-age interaction. RVI-ASD showed significant positive correlation with paternal age, while correlation with maternal age was non-significant. Baseline and follow-up RVI-ASD were negatively correlated with cognitive measures including total, fluid and crystallized intelligence (r=-0.09 to -0.11, p<10-6). RVI-ASD scores tracked with core autism phenotypes including poor eye contact and rigid routines (p < .01). In a sub-sample of children with symptoms of autism (N=20), we found that earlier age of symptom onset was strongly correlated with higher White Matter RVI (r = -0.61), linking early behavioral emergence to the neuroanatomical signature. Longitudinal changes in subcortical RVI-ASD are significantly correlated with change in social functioning. The RVI-ASD is a neuroimaging-based biomarker linked to stable and reproducible brain patterns in autism. RVI-ASD allows researchers to study associations with factors associated with the likelihood for autism and cognition across large and inclusive non-clinical samples, thus moving beyond simple case-control models to understand the biological pathways of autism.

neuroscience↗

bioRxiv2026

cGAS bends unpaired DNA to form an unconventional structure that hyperactivates the innate immune response

cGAS is a pattern-recognition receptor for dsDNA and forms 2cGAS:2DNA dimers followed by oligomerization into phase-separated condensates when fully-complementary DNA is studied. However, many DNAs are not fully complementary. We report that DNA with unpaired regions such as those found during transcription, recombination or replication (designated as bubble-DNA, Bu-DNA) causes cGAS hyper-activation. Hyperactivation is observed by Bu-DNA embedded in linear DNA, circular DNA, plasmid DNA and mitochondria DNA. Bu-DNA binds significantly more tightly to the cGAS catalytic domain than paired-DNA but suppresses condensation. Cryo-EM and single-molecule FRET reveal that cGAS forms 2cGAS:1DNA complexes by bending Bu-DNA into a V-shape using the unpaired region as a hinge, limiting its oligomeric state. This uncovers a novel mode of cGAS activation attributed to pattern diversity within pattern ligands.

immunology↗

bioRxiv2026

Mapping Tumor-Microenvironment dependencies with TMEformer: A spatial foundation framework enabling in silico perturbation

Despite the fundamental role of spatial context in driving tumor progression, most current computational models for virtual perturbation have largely overlooked its importance. Here, we introduce TMEformer, a tumor microenvironment-aware deep learning framework that leverages high-resolution spatial transcriptomics to jointly model intrinsic tumor cell programs and local microenvironmental signals by explicitly incorporating spatial architecture. Validated across diverse tumor spatial transcriptomic cohorts, TMEformer enables virtual perturbations that capture functional dependencies within local cellular ecosystems. Despite being trained on cancer-specific spatial datasets, TMEformer outperforms baseline models pretrained on large-scale corpora in capturing key tumor transitions, including lineage plasticity and the emergence of therapy resistance. Systematic perturbation analyses prioritize tumor-intrinsic transcription factors and TME-derived ligands that drive disease progression, recovering established regulators and revealing novel candidates. Furthermore, TME- derived embeddings improve the spatial stratification of tumor cells and align more closely with pathological architecture. Together, TMEformer establishes a general framework for modeling tumors as spatially coupled, perturbable ecosystems. An online TMEformer web platform for perturbation prediction, analysis, and visualization is available at http://www.pradcellatlas.com.

bioinformatics↗

bioRxiv2026

Structure of the human HIRA histone chaperone with a nucleosome suggests a stepwise nucleosome assembly mechanism

The restoration of chromatin in the wake of a DNA or RNA polymerase is essential to maintain the integrity of eukaryotic genomes. Human HIRA is a 1.8-megadalton, three-subunit histone chaperone that mediates all replication-independent deposition of the histone variant H3.3 at active chromatin regions1-6. Disruption of HIRA perturbs active-chromatin organization and has wide-ranging consequences for development, cellular senescence, and genome integrity7-11. Despite its central biological role in reassembling nucleosomes post-transcription, the structure of native human HIRA and the mechanism by which it organizes histones and DNA during nucleosome assembly remain unknown. In particular, the function of the largest HIRA subunit CABIN1 is enigmatic. Here, we show that HIRA is not simply a passive histone hand-off factor but remains engaged across multiple stages of nucleosome assembly, including a close interaction with the nucleosome. Cryo-EM structures reveal that HIRA forms an extended arch-like structure that binds the nucleosome primarily through extensive CABIN1 contacts with histones, histone tails, nucleosomal DNA, and linker DNA, during the final stage of nucleosome assembly. Together, our results suggest a testable mechanism for HIRA-mediated nucleosome assembly and product release and provide the basis for elucidating the molecular details of this fundamental biological process.

biochemistry↗

bioRxiv2026

Targeting Microbial Bile Salt Hydrolase Reprograms Bile Acid Metabolism and Ameliorates Metabolic Dysfunction-Associated Steatohepatitis in Mice

Microbial bile salt hydrolase (BSH) plays a central role in shaping bile acid composition and gut-liver metabolic signaling, yet its therapeutic potential in metabolic dysfunction-associated steatohepatitis (MASH) remains incompletely defined. Here, we evaluated the efficacy of the non-absorbable BSH inhibitor GR-7 in a diet induced mouse model of steatohepatitis using early and late intervention strategies with different dosing regimens. GR-7 reduced food intake and exerted stage- and dose-dependent therapeutic effects, with early intervention robustly suppressing hepatic fibrosis even at low dose, whereas late-stage administration of high-dose GR-7 markedly reduced hepatic steatosis and inflammation, as evidenced by decreased liver weight, hepatic triglyceride and cholesterol levels, and plasma ALT. Although late intervention did not result in statistically significant histological reversal of fibrosis, a trend toward improvement was observed, together with suppression of fibrogenic gene expression, suggesting that prolonged treatment may further enhance antifibrotic efficacy. Mechanistically, GR-7 effectively inhibited microbial BSH activity in vivo, leading to reduced cecal unconjugated primary and secondary bile acids--including deoxycholic acid and lithocholic acid, which was associated with improved gut barrier integrity and reduced hepatic inflammation. In parallel, BSH inhibition reprogrammed hepatic bile acid metabolism toward activation of the alternative CYP27A1-mediated synthesis pathway, accompanied by reduced food intake, thereby contributing to improved hepatic lipid accumulation. Furthermore, late-stage high-dose treatment selectively remodeled the hepatic immune landscape rather than fully restoring homeostasis, highlighting immune recalibration as a key component of therapeutic response. Together, these findings identify microbial BSH inhibition as a promising microbiome-targeted therapeutic strategy for MASH. HighlightsO_LIThe non-absorbable BSH inhibitor GR-7 improves steatosis, inflammation, and fibrosis in of Western diet-induced steatohepatitis model in mice in a dose-dependent manner. C_LIO_LIGR-7 reduces food intake and body weight gain. C_LIO_LIGR-7 reduces cytotoxic secondary bile acids, including DCA and LCA. C_LIO_LIGR-7 reprograms hepatic bile acid metabolism and immune responses. C_LI

pharmacology and toxicology↗

bioRxiv2026

Neural Substrates of Duration Serial Dependence: Opposing Modulation in Basal Ganglia and Posterior Medial Cortex

Perceptual judgments are systematically biased toward recent experience, a phenomenon called serial dependence. The brain circuits underlying this effect in time perception remain unknown. Using fMRI, we identified the neural substrates of duration serial dependence and tested whether they overlap with those for motion direction. Twenty-seven participants reproduced either the duration or direction of a visual stimulus; a retro-cue revealed which feature to report 3-s after stimulus onset, separating memory-trace signals from task-switching signals. Duration serial dependence was robust and roughly four times stronger on task-repeat than task-switch trials; direction serial dependence was near zero. The previous duration produced opposing responses in the bilateral putamen (suppression) and in the posterior cingulate cortex and precuneus (reinstatement); no analogous signal emerged for direction. Task switching from direction to time task engaged the inferior frontal junction and the dorsolateral prefrontal cortex, which carried no serial dependence signal. The anatomical dissociation resolves the behavioral pattern: the duration trace is encoded in the putamen and posterior medial cortex on every trial, but task-set reconfiguration by the frontoparietal network gates its read-out into behavior, explaining the task repetition effect. These results reveal a dissociation between two functionally distinct circuits: the putamen and posterior cingulate/precuneus carry signals from the previous trial, whereas frontoparietal regions support task switching but carry little serial dependence information.

neuroscience↗

bioRxiv2026

Massilia varians P2-4, A Potential Biocontrol Agent against Pathogenic Pseudomonas aeruginosa in Eriocheir sinensis

Pseudomonas aeruginosa is a clinically significant bacterial pathogen that poses a serious threat to aquaculture. However, there are limited information on Massilia isolates against pathogenic P. aeruginosa in aquaculture. In the present study, a facultative predator, M. varians isolate P2-4, was isolated from aquaculture sediment using Chinese mitten crab Eriocheir sinensis-pathogenic P. aeruginosa as the prey bacterium, and its genomic feature, bacteriolysis-related genes, safety, bacteriolytic spectrum, and in vitro and in vivo antibacterial effects against pathogenic P. aeruginosa in E. sinensis were further characterized. Isolate P2-4 consisted of one chromosome and one plasmid (with a total of 75 tRNAs, 7 5S rRNAs, 7 16S rRNAs, 7 23S rRNAs, 34 sRNAs, 5,238 coding genes, 20 genomic islands, 1 prophage, 23 insertion sequences, and 102 repeat sequences), and harbored 19 bacteriolysis-related genes (pilA, pilB, pilC, pilD, pilF, pilG, pilH, pilM, pilO, pilP, pilQ, pilS, pilR, pilT, mltA, mltB, mltC, mltD, and dacB) associated with cellular motility and cell wall lysis. In addition, the isolate carried no virulence genes, was unable to produce haemolysin, hydrogen sulfide, nitrite and ammonia, and avirulent in E. sinensis with a 7-day acute intraperitoneal LD50 value of above 5.0 x 108 CFU/mL. Furthermore, the isolate possessed a wide bacteriolytic spectrum against pathogenic Shewanella algae, Aeromonas caviae, A. hydrophila, and Photobacterium damselae besides P. aeruginosa, exhibited bacteriolysis rates of 99.35% to 99.99% towards the pathogenic P. aeruginosa at 1.0x103 to 1.0x10{square} CFU/mL, and displayed relative percentage survivals of 42.31% to 73.08% against P. aeruginosa infection in E. sinensis at doses of 6.0 x 103 to 6.0 x 105 CFU/g diet. To our knowledge, this study for the first time demonstrates a M. varians strain as a potential biocontrol agent against pathogenic P. aeruginosa in aquaculture.

microbiology↗

bioRxiv2026

Engineering a pacemaker-driven human mini-heart guided by spatial multi-omics of sinoatrial node development

The human sinoatrial node (SAN) functions as the primary pacemaker of the heart and coordinates the hierarchical electrical activity that drives cardiac contraction. However, experimental systems capable of reconstructing pacemaker-driven cardiac organization in human tissues remain limited. Here we integrate spatial multi-omics of the human fetal SAN with stem-cell engineering to generate pacemaker organoids ("Sinoids") and assemble them into a pacemaker-driven human mini-heart composed of sinoatrial, atrial and ventricular cardiac modules. High-resolution spatial transcriptomics and single-nucleus multi-omic analyses of human fetal SAN tissues identify regulatory pathways guiding pacemaker lineage specification, which we leverage to engineer human pluripotent stem cell-derived SAN organoids with robust pacemaker identity and electrophysiological activity. When integrated with atrial and ventricular cardioids, Sinoids initiate and coordinate electrical activation across assembled cardiac tissues, establishing directional propagation of electrophysiological signals within structured mini-heart organoids. Combining AI-guided perturbation modeling with functional validation further identifies conserved regulatory pathways controlling pacemaker specification and regionalization, including YAP-TEAD and NRG-ERBB signaling. Together, these results establish a multi-omic-guided strategy for engineering pacemaker tissues and reconstructing cardiac conduction hierarchy in vitro. The pacemaker-driven mini-heart platform provides a modular human cardiac system for studying pacemaker biology, modeling arrhythmia mechanisms and enabling electrophysiological drug discovery.

bioinformatics↗

bioRxiv2026

Environmental Volatility Shifts Visual Search from Capture to Caution

Real-world distractors occur in environments whose states change at different rates. We asked whether such volatility alters early attentional gating or instead changes the criterion for committing to a response. Observers performed an additional-singleton search task with concurrent eye tracking while distractor presence followed high- or low-volatility sequences, with overall distractor prevalence held constant. Trial-pooled oculomotor capture was higher under high volatility, a pattern that appears to indicate altered filtering. That inference did not survive repetition-aware analysis: once the same-location run position was matched, capture did not detectably differ across volatility regimes. The pooled capture effect was therefore consistent with a structural consequence of the volatility manipulation, which enriched high-volatility blocks with early-run positions where capture is intrinsically high. The positive volatility signature appeared on distractor-absent trials, where high-volatility blocks were associated with longer target latency, more fixations, longer final-target dwell, and fewer errors. Same-location repetition learning showed no detectable difference in slope across regimes. A hierarchical drift-diffusion model (DDM) and a complementary volatility Kalman-filter (VKF) dynamic-state comparison indicated that manual responses were better described by architectures that allow both boundary-related and drift-related components than by a boundary-only account. Volatility, therefore, did not show detectable evidence of impairing the local gating rule; instead, the converging evidence points to a post-selective verification/caution profile, consistent with a precision-weighted read-out of environmental uncertainty.

neuroscience↗

bioRxiv2026

miR-128 Regulates Hypertensive Vascular Remodeling via PPAR-γ

OBJECTIVEThis study aimed to explore the role and underlying mechanism of microRNA-128 (miR-128) in regulating vascular remodeling in spontaneously hypertensive rats (SHRs), focusing on its targeting of peroxisome proliferator-activated receptor {gamma} (PPAR-{gamma}) and modulation of the Toll-like receptor 4/nuclear factor-{kappa}B (TLR4/NF-{kappa}B) inflammatory pathway. METHODSAll experimental procedures were approved by the Animal Care and Use Committee of Fujian Medical University. In vivo, ten-week-old male SHRs were randomly assigned to three groups: renal denervation (RDN, n=6), sacubitril/valsartan (Sac/Val, n=6), and Sham (n=6). Age-matched Wistar-Kyoto (WKY) rats served as normotensive controls (n=6).Eight weeks after intervention, mesenteric arteries were harvested for histological, functional, and molecular analyses. Serum miR-128 levels were measured by quantitative real-time polymerase chain reaction (qRT-PCR). The expression levels of key proteins in the vascular wall were assessed via immunofluorescence (IF), immunohistochemistry (IHC), and Western blotting (WB). Bioinformatics analysis and RNA sequencing (RNA-seq) were employed to identify core genes and signaling pathways associated with hypertension-induced pathological inflammation. RESULTSIn vivo, in the SHR sham-operated group, elevated blood pressure, severe vascular remodeling, and impaired vasodilatory function were observed, accompanied by downregulated miR-128 expression and upregulated TLR4/NF-{kappa}B signaling activity (all p < 0.0001).RDN postoperative, miR-128 expression was significantly restored, which in turn inhibited the TLR4/NF-{kappa}B pathway, reduced the production of pro-inflammatory cytokines (including IL-1{beta}, IL-6, and TNF-), and ameliorated vascular dilation dysfunction in SHRs (all p < 0.0001). Mechanistically, miR-128 negatively regulated the TLR4/NF-{kappa}B signaling pathway while upregulating the expression of PPAR-{gamma} (p < 0.05). CONCLUSIONRDN not only exerts a hypotensive effect but also improves hypertensive vascular remodeling. miR-128 inhibits excessive inflammation in vascular smooth muscle cells and alleviates vascular remodeling in SHRs via the PPAR-{gamma}/TLR4/NF-{kappa}B axis. These findings identify miR-128 as a potential therapeutic target for RDN in the treatment of hypertension, providing a novel regulatory strategy for the precision management of cardiovascular diseases.

bioinformatics↗

bioRxiv2026

AEGIS reveals epitope- and clone-resolved convergence of CNS B and T cell autoreactivity in ROHHAD

Autoimmune diseases arise when B and T lymphocytes lose tolerance to self. Yet in most disorders, the underlying molecular determinants, including autoantibodies, epitopes and lymphocyte clones that drive tissue injury remain undefined. Rapid-onset obesity with hypothalamic dysfunction, hypoventilation and autonomic dysregulation (ROHHAD) is a rare and often fatal pediatric neuroendocrine syndrome with strong evidence of antigen-driven paraneoplastic autoimmunity, including association with the intracellular autoantigen ZSCAN1. However, the effector immune circuit and the epitope-level determinants operating within the hypothalamus and brainstem have remained unknown. To address this challenge in ROHHAD and more broadly in autoimmune disease, we developed the Autoimmune Epitope and immunoGlobulin/Immune-receptor identification System (AEGIS), an integrated framework that links immune repertoires to their cognate self-epitopes. AEGIS combines B cell and T cell receptor profiling from sites of tissue injury with high-resolution epitope mapping, direct sequencing of antigen-specific autoantibodies, in silico antibody-antigen folding, selection, and T cell antigen discovery. Applied to a deeply phenotyped child with ROHHAD, AEGIS revealed a compartmentalized, clonally restricted immune response in which brain-deposited IgG and expanded cerebrospinal fluid B cell and CD4 T cell clonotypes converged on shared ZSCAN1 epitopes, resolved to minimal determinants and peptide-MHC ligands. These findings provide a clone- and epitope-linked mechanistic map of ROHHAD autoimmunity and establish a generalizable framework for identifying candidate pathogenic clones and antigens across diverse autoimmune diseases.

immunology↗

bioRxiv2026

Bacteria rewire fungal antimicrobial gene expression in microbial arms races

Microbial communities are shaped by antagonistic interactions, yet how competitors evade antimicrobial strategies in a community context remains unclear. Fungi deploy antimicrobial proteins (AMPs) to suppress bacterial niche competitors, but how bacteria counteract these defences is unknown. Here we show that soil microbiota suppress fungal AMP expression, revealing a previously unrecognized mode of intermicrobial interference. In the soil-borne pathogen Verticillium dahliae, AMP genes are highly conserved yet broadly repressed in the presence of natural microbial communities. This repression extends across diverse fungi, suggesting a widespread phenomenon. Correlation analyses of AMP expression and community compositions linked the bacterial genera Leptothrix and Devosia to repression of the V. dahliae AMPs VdL1 and VdD1, respectively. Both AMPs antagonize these bacterial taxa, and microbiota-mediated repression of VdL1 reduces its contribution to fungal soil colonization. Together, our findings reveal that bacteria can evade fungal antagonism by suppressing AMP deployment, highlighting community-level control of microbial competition.

microbiology↗

bioRxiv2026

UTag, a cysteine-free thermostable tagging system for tracking single mRNA translation live

Spatiotemporal regulation of mRNA translation is central to gene expression. Over the past decade, translation has become directly observable in live cells at single-mRNA resolution by tagging nascent chains with tandem arrays of short epitope tags recognized by genetically encodable fluorescent intracellular antibodies (intrabodies). While this technology has revolutionized our understanding of translation regulation, the current toolbox of tagging systems remains limited. Here, we developed a novel and tight-binding intrabody against a short (11-amino acid) HIV protease epitope (named UTag). To ensure robust intracellular folding of the anti-UTag intrabody, we further engineered a cysteine-free variant that folds and functions independently of disulfide-bond formation, as validated by X-ray crystallography. The cysteine-free anti-UTag intrabody retains high binding affinity comparable to the parental intrabody while exhibiting significantly improved thermostability ([~]80 {degrees}C). Importantly, the cysteine-free UTag system enables real-time tracking of single-mRNA translation in live cells with performance on par with the parental UTag system as well as the established SunTag and ALFA-tag. Collectively, these results demonstrate that the newly developed UTag system expands the toolbox for live-cell translation tracking and provides complementary tools for multiplexed applications.

biophysics↗

bioRxiv2026

The Par complex regulates apical-basal cell polarity through modulation of FAK signaling homeostasis

Cell polarity complexes are essential for embryogenesis, but their regulatory mechanisms during early developmental transitions remain incompletely understood. Here, we individually deleted the Crumbs, Par, and Scrib polarity complexes in mouse embryonic stem cells (mESCs). While loss of any single complex did not affect pluripotency or proliferation, deletion of Par complex disrupted the naive-to-primed transition and impaired subsequent differentiation, particularly lumen formation in neural tube organoids. Mechanistically, Par complex deficiency led to hyperphosphorylation of focal adhesion kinase (FAK) at the primed stage, driving a morphological shift from flat monolayer clusters to dome-shaped colonies. FAK inhibition rescued the aberrant morphology. Upstream, Par complex loss increased AKT phosphorylation, which remodeled extracellular matrix (ECM) and regulated integrin signaling via FURIN-LEFTY, ultimately modulating FAK activity. In addition, conditioned medium from wild-type cells partially rescued differentiation defects in Par knockout cells in a LEFTY-dependent manner. These phenotypes were consistently observed in naive-to-primed transition, neural stem cell differentiation, embryoid body formation, teratoma assays, and neural tube organoid differentiation. Together, these findings establish a Par complex-AKT-FURIN-LEFTY-ECM-integrin-FAK signaling cascade that links apical-basal polarity to early lineage specification and morphogenesis, providing a mechanistic framework for how polarity cues are translated into developmental outcomes. Graphical Abstract O_FIG O_LINKSMALLFIG WIDTH=200 HEIGHT=200 SRC="FIGDIR/small/722465v1_ufig1.gif" ALT="Figure 1"> View larger version (64K): org.highwire.dtl.DTLVardef@179e0adorg.highwire.dtl.DTLVardef@8d83bdorg.highwire.dtl.DTLVardef@c425fforg.highwire.dtl.DTLVardef@19a8ee5_HPS_FORMAT_FIGEXP M_FIG C_FIG Significance StatementThis study elucidates the molecular mechanism by which the Par complex regulates the establishment of cell polarity. The authors demonstrate that the Par complex promotes the expression of the protein convertase FURIN via AKT signaling, thereby enhancing the maturation and secretion of LEFTY protein. This process remodels the ECM and modulates integrin signaling, ultimately regulating FAK activity and controlling the establishment of cell polarity. These findings reveal how polarity cues govern early lineage specification and morphogenesis, with implications across multiple developmental contexts.

developmental biology↗

bioRxiv2026

A low-dimensional transcriptional code enables decoding of BMP/TGFβ signaling from single-cell transcriptomes

Cells interpret complex extracellular environments through signaling pathways that compress diverse ligand inputs into a limited set of intracellular mediators. How this information is encoded in transcriptional responses, and whether it can be decoded to infer the original signaling environment, remains unclear. Here, we systematically map BMP/TGF{beta} transcriptional responses using single-cell RNA sequencing across 48 ligand-concentration conditions, generating a multi-ligand dose-response atlas. We find that, for each ligand, transcriptional responses collapse onto a one-dimensional, highly coordinated program in which concentration modulates response amplitude without altering gene identity. Across ligands, we find a small number of distinct, ligand-dependent transcriptional programs. We leverage this structured encoding to define a quantitative perception score that captures pathway activity at single-cell resolution. Finally, we train a machine learning model to decode extracellular ligand concentrations from single-cell transcriptomes. The model further generalizes to in vivo intestinal epithelium, where it reconstructs the spatial BMP gradient profile. Together, these results reveal a low-dimensional transcriptional code in BMP/TGF{beta} signaling that constrains cellular responses and enables quantitative inference of extracellular signaling environments from gene expression.

systems biology↗

bioRxiv2026

Forest expansion and open vegetation responses during the past 14 ka at Zhagaer Co on the eastern Tibetan Plateau

Forest expansion and retreat are key drivers of alpine ecosystem dynamics, yet it remains unclear how forest expansion shapes alpine plant communities and whether alpine assemblages that re-emerge after forest decline resemble those of the Late Glacial. Here, we integrate sedimentary ancient DNA (sedaDNA) and pollen records from Zhagaer Co on the eastern Tibetan Plateau to reconstruct vegetation changes over the past [~]14 ka. Using Picea abundance as a proxy for forest cover, we apply constrained ordination of sedaDNA-inferred plant communities to identify plant taxa as "winners" and "losers" of forest expansion and compare the composition of loser taxa between the late Glacial and late Holocene cold-open phases. Our results show that forest expansion during the early to mid-Holocene favoured woody and forest-margin taxa (e.g. Rhododendron, Salicaceae), while suppressing alpine forbs and graminoids (e.g. Carex, Thalictrum), consistent with patterns expected under ecological filtering. However, late Holocene reopening did not fully restore a late Glacial-like alpine community; instead, it was characterised by a stronger contribution of alpine meadow and shrub taxa. This difference may reflect contrasting environmental backgrounds, including higher atmospheric CO2 levels during the late Holocene, ecological legacies of prior forest expansion, and increasing human influence. These findings suggest that forest expansion may lead to long-term restructuring of alpine plant assemblages, and that late Holocene cooling did not simply restore late Glacial alpine communities but instead produced a distinct alpine ecosystem state. Together, these results highlight the long-term legacy of treeline dynamics in shaping alpine ecosystem trajectories.

ecology↗

bioRxiv2026

A Decade of Deep Learning-based Biomedical Image Segmentation

Biomedical image segmentation is a fundamental problem in computational biomedicine that aims to precisely delineate anatomical and biological structures, tissue types, or pathological regions in biomedical images. Accurate segmentation is essential for interpretation, decision-making, and quantitative analysis across a wide range of biological and medical applications. Over the past decade, the field has undergone a profound paradigm shift, evolving from task-specific specialist models to universal foundation models. This review provides an in-depth analysis of the evolution, tracing how the limitations of local discriminative learning drove the transition toward transformer-based global modeling, and large-scale generative pre-training. To help navigate the diverse landscape of interaction paradigms, we introduce the first systematic taxonomy of promptable biomedical image segmentation, categorizing existing methods into six distinct types, enabling users to intuitively select appropriate prompting strategies based on visual demonstrations and quickly pinpoint relevant literature (Prompt Type Visualization). Beyond model architectures, we discuss parallel advancements in dataset development, evaluation protocols, and application-specific adaptations across radiology, pathology, and biology. Integrating these powerful foundation models with rigorous domain-specific adaptation has great potential to improve patient outcomes and healthcare efficiency. Finally, we highlight key challenges in trustworthiness and clinical integration that must be overcome to realize the potential of the next generation of biological and medical generalists.

bioengineering↗

bioRxiv2026

Predicting Supramolecular Self-Assembly of Peptide Structures with AlphaFold3

Self-assembled peptide-based nanostructures have diverse applications in the pharmaceutical and materials fields, but accurately predicting their self-assembly behavior without time-intensive organic synthesis and characterization remains a significant challenge. Here, we assess the effectiveness of AlphaFold3 (AF3), a deep learning model for protein structure prediction, in modeling peptide-based nanostructures and the interactions driving supramolecular self-assembly. We designed amphiphilic peptides composed of alternating hydrophobic residues (valine, leucine, isoleucine, phenylalanine) and hydrophilic residues (glutamic acid), varying both sequence length and residue order. Using AF3's multimer mode, we modeled assemblies with copy numbers ranging from 10 to 1000, generating diverse morphologies such as micelles and nanotubes. We qualitatively analyzed hydrophobic regions, secondary structures, and intermolecular interactions, while also calculating radii of gyration, packing scores, and aspect ratios using PyRosetta. Our results indicate that AF3 predicts morphologies consistent with hydrophobic driving forces and steric constraints. Increased hydrophobicity correlates with smaller radii of gyration, while higher copy numbers correspond to smaller aspect ratios (more compact structures). Longer hydrophobic segments lead to disordered structures, whereas longer hydrophilic segments promote organization. While AF3 captures systemic trends consistent with biophysical principles, comparisons to literature reveal discrepancies driven by charge effects and secondary structure bias, including an overemphasis on helical propensity (e.g., alanine-rich sequences) and sensitivity to terminal charge repulsion. Additionally, since AF3 is predisposed to predict a single assembled entity rather than higher-order assemblies such as multiple micelles or fibers, finding the optimal copy number for the best prediction requires system-specific iteration. These limitations highlight the need for complementary approaches with controlled chemical potential and environmental conditions, though qualitative agreement with experimental trends in morphology and compactness supports AF3's utility for initial structure generation. Our findings highlight AF3's potential as a user-friendly first pass screening tool for peptide design, enabling rapid generation and prioritization of candidate assemblies for subsequent evaluation using coarse-grained or all-atom molecular simulations toward the efficient development of functional self-assembled peptide nanomaterials.

bioengineering↗