Biology subjects

Ding, W.

Publications and source records attributed to Ding, W..

At least 19 recordsLinked to original sources

bioRxiv2026

Decision benchmarking reveals when protein foundation models add experimental value

Protein engineering is shifting from model-based variant scoring to deciding what to measure next. This exposes a limitation of fixed-dataset benchmarks, which rank predictors without showing when their advantage survives as evidence accumulates. We therefore introduce decision benchmarking across 217 ProteinGym assays, 2.47 million variants and 95 predictors to evaluate whether models and measurements improve decisions at matched cost. Zero-shot accuracy explained less than one-fifth of the variation in discovery after calibration. Calibrated priors exceeded budget-matched additivity at 20-100 measurements, but the advantage narrowed as component measurements accumulated. Measured additivity then overtook zero-shot models after 50-100 measurements for top-variant discovery and around 500 for global ranking. The same principle showed that double-mutant measurements improved prediction in models using mutation identity, whereas prediction-guided acquisition improved discovery. Evidence-based routing improved strategy selection. Decision benchmarking thus identifies which strategy adds value given the information already available and the decision that follows.

bioengineering↗

bioRxiv2026

A claim-centred atlas distinguishes biological corroboration from analytical depth in plant stress transcriptomes

Plant stress transcriptomes are reanalysed far more often than independently replicated, obscuring whether apparent consensus reflects biological corroboration or repeated analysis of the same experiments. Here we present PLEASE (PLant Evidence Atlas for StrEss), a claim-centred atlas that treats the independent experiment, rather than the analysis, as the unit of evidence. PLEASE resolves 88,789 analytical records into 62 dependency-resolved study families. A calibrated core of abiotic and biotic stress programmes persists after dependency collapse. Yet support erodes sharply as claims become more specific. Across stress, organ and cell-lineage contexts, independent-match coverage falls from 1.00 to 0.26 while directional agreement remains near 0.68. Three quarters of sampled contexts contain only one study family. Thus, missing comparable experiments, not pervasive disagreement, is the main barrier to context-specific corroboration. PLEASE separates analytical depth from biological corroboration and pinpoints where independent experiments are most needed.

plant biology↗

bioRxiv2026

Natural product-derived azaphilones are selective scaffolds for covalent lysine liganding in live cells

Lysine-directed covalent liganding offers a promising strategy to expand the covalently ligandable proteome beyond cysteine. However, broadly applicable and selective lysine-reactive scaffolds remain limited. Here, we designed SCL-alkyne, a simplified azaphilone probe derived from the natural product sclerotiorin (SCL), for the global profiling of lysine reactivity in bacteria. SCL-alkyne displayed high intrinsic selectivity for lysine in vitro and in live Escherichia coli cells. In situ proteome profiling revealed that 84% of engaged sites belong to lysine residues and the protein N-terminus. Reactivity mapping provided an inventory of highly reactive sites showing that only a limited subset of reactive lysines is located in functional regions of proteins. Analysis of proteins enriched via the SCL-alkyne probe unraveled bacterial chemotaxis as a major target pathway which was confirmed by an impaired chemotactic response upon treatment of bacteria with SCL-alkyne. Competitive profiling of three structurally related azaphilone natural products identified 60 protein targets belonging to pathways related to iron-sulfur cluster assembly and antioxidant function. This study not only highlights the diverse modes-of-action of natural product azaphilones, but also establishes its core scaffold as effective lysine-reactive chemotype for identifying ligandable lysines.

biochemistry↗

bioRxiv2026

Understanding Strategic Motor Learning as a Process of Hypothesis Testing

Multiple learning processes contribute to successful goal-directed actions under changing physiological states, biomechanical constraints, and environmental contexts. Among these, explicit strategies enable us to discover new movement patterns when existing ones no longer achieve the desired outcome. Yet, how strategies are discovered during motor learning remains unknown. To address this, we developed a novel behavioral paradigm that isolates strategy discovery in response to a range of visuomotor perturbations. This approach revealed that strategy discovery unfolds through an initial period of systematic exploration across multiple candidate strategies, followed by an 'aha' moment in which behavior converges on a stable solution. To account for these dynamics, we developed a computational model based on hypothesis testing in which learners generate and evaluate visuomotor rules to counteract the perturbation. This hypothesis testing model outperformed a range of alternative accounts, including gradual error reduction, win-stay lose-shift learning, and sudden one-shot insight. Together, these findings identify hypothesis testing as a computational mechanism by which humans discover new strategies during motor learning.

neuroscience↗

bioRxiv2026

FLRT cell adhesion molecules counteract protocadherin-10-dependent cell segregation during cortical neuron migration

The processes that underlie the folding of the cerebral cortex of large mammals and the evolution of the smooth cortex of modern rodents are incompletely understood. Recent evidence highlighted adhesion-controlled neuronal migration as an important mechanism. Genetic deletion in mice of FLRT1 and FLRT3 cell-adhesion proteins perturbed cortical neuron migration, lead to region-specific neuronal clustering, and introduced cortical folds into the normally smooth mouse cortex, suggesting that Increased expression of FLRT protein may have contributed to cortex smoothening. Here, we identify protocadherin-10 (PCDH10) as a key regulator of neuronal clustering in this context. Using single-cell transcriptomics and spatial proteomics, we describe a cortical neuron subtype in which loss of FLRT1/3 triggers local upregulation of PCDH10 in a specific cortical region. Elevated PCDH10 drives homophilic adhesion and the formation of cell patches that segregate from FLR1/3-deficient migrating neurons. This segregation leads to discrete clustering of Flrt1/3-mutant neurons and correlates with the emergence of cortical folds. Consistent with an evolutionary role, comparative expression analysis showed reciprocal expression of FLRT3 and PCDH10, with high FLRT3 and low PCDH10 expression in the smooth mouse cortex, and the opposite pattern in the developing human cortex. Together, our findings reveal how the antagonism between different classes of cell adhesion molecules affects cortical neuron migration patterns, contributes to region-specific cortical folding and may have promoted evolutionary smoothening of the rodent cortex.

neuroscience↗

bioRxiv2026

Ergothioneine, alone or combined with vitamin K2, vitamin D3 and magnesium L-threonate, attenuates bone turnover, inflammatory and oxidative disturbances in ovariectomized mice

PurposeEstrogen deficiency drives bone loss through a complex interplay of endocrine, oxidative, inflammatory, and bone-remodeling disturbances. Ergothioneine (EGT) is a diet-derived thiol/thione antioxidant, yet its effects on the estrogen-deficient skeleton remain unknown. This study evaluated whether EGT-alone or combined with vitamin K2, vitamin D3, and magnesium L-threonate-attenuates the skeletal and systemic consequences of ovariectomy (OVX) in mice. MethodsForty-eight female C57BL/6J mice underwent either sham surgery or OVX. For 12 weeks, the mice received a daily oral gavage of one of the following: vehicle, alendronate (1.53 mg/kg), EGT (30 mg/kg/day), EGT with vitamin K2 (40 {micro}g/kg/day) and vitamin D3 (500 IU/kg/day), or EGT with vitamin K2, magnesium L-threonate (350 mg/kg/day), and vitamin D3 (n = 5-6 per group). Outcomes evaluated included the uterine index, tibial micro-computed tomography (micro-CT), distal-femoral histology, serum bone turnover markers (CTX-I, PINP, osteocalcin), sex hormones, TNF-, IL-6, SOD, and MDA. ResultsOVX significantly lowered the uterine index and induced tibial trabecular deterioration. This was accompanied by increased CTX-I, TNF-, IL-6, and MDA, alongside decreased PINP, osteocalcin, and SOD (all P < 0.01 vs. sham). Alendronate successfully restored tibial micro-CT bone-volume fraction (BV/TV) and trabecular number (P < 0.01 vs. OVX). While the EGT-based regimens did not significantly restore tibial micro-CT BV/TV, trabecular thickness, or trabecular number (all P > 0.05 vs. OVX), they significantly increased the trabecular area in distal-femoral histology (14.2-15.0% across regimens vs. 7.6% in OVX; P < 0.05). Furthermore, EGT treatments lowered CTX-I, TNF-, IL-6, and MDA, increased SOD, and partially restored PINP and osteocalcin (P < 0.05-0.01 vs. OVX). Apparent increases in serum estradiol were assay-dependent and are considered exploratory. ConclusionEGT-based nutritional regimens improved the systemic oxidative, inflammatory, and bone-turnover environments associated with estrogen-deficient bone loss and successfully preserved the distal-femoral trabecular area. However, tibial three-dimensional microarchitecture was not restored on micro-CT. Given that histological and micro-CT endpoints were evaluated at different skeletal sites, structural interpretations should be made with caution. These findings support the further evaluation of EGT as a dietary adjunct, warranting subsequent mechanistic and dose-optimization studies.

pharmacology and toxicology↗

bioRxiv2026

Aging promotes inflammation and steatosis in alcohol-associated liver disease in mice

Background and AimsAs the older population (aged 65 years and older) continues to expand and more people drink alcohol, aging has been linked to the development of alcohol-associated liver disease (ALD) and to worse disease outcomes. The aim of this study was to explore the mechanisms by which advanced age and alcohol exacerbate alcohol-induced liver injury. MethodsTwo-to-three-month-old and twenty-to-twenty-two-month-old male C57BL/6N mice were subjected to chronic-on-binge alcohol feeding (the Gao-binge model). Serum alanine aminotransferase, aspartate aminotransferase, and triglyceride content were determined using biochemical assays. The levels of lipogenesis, fatty acid-metabolizing proteins, inflammatory markers, mitochondrial and autophagy-related proteins, and senescence-associated proteins were determined by immunoblotting, immunohistochemistry, and real-time quantitative polymerase chain reaction (RT-qPCR). Liver tissues were also subjected to RNA sequencing and metabolomics analyses. Proteomics analysis was performed on serum samples. Tail-vein adenovirus-TFEB was injected to overexpress hepatic TFEB in 22-month-old C57BL/6N male mice, followed by Gao-binge alcohol feeding. ResultsHepatic triglyceride content was significantly increased in aged, alcohol-fed mice, whereas serum ALT and AST levels remained relatively similar between alcohol-fed young and aged mice. Gao-binge alcohol increased the hepatic levels of diacylglycerol and acyl-carnitine species in both aged and young livers. RNA sequencing, proteomic analysis, and serum cytokine array analysis showed that inflammatory cytokines, including Ccr2, Cxcl1, and CCL6, and pro-inflammatory antibody fragments were increased in aged, alcohol-fed mice. Increased gene and protein expression of the senescent markers p21 and p27, along with increased senescent-associated (SA) {beta}-galactosidase activity in ethanol-fed aged mice compared to young mice. Gene and protein expression of TFEB was downregulated in ethanol-fed young and aged animals, along with decreased levels of lysosomal ATPases and hepatic dipeptide content. Overexpression of TFEB in the livers of aged, Gao-binge-fed mice was associated with reduced Ly6G-positive cells, reduced protein levels of the innate immune mediators cGAS, IRF-7, IRF3, and NLRP3, and reduced caspase-1 activity as well as serum ALT levels. ConclusionsOur findings indicate that advanced age perpetuates the detrimental effects of excessive alcohol consumption on various homeostatic processes and promotes steatosis and inflammation in the liver. Modulations in hepatic TFEB could be effective in mitigating pro-inflammatory signaling that occurs due to the synergistic effect of both heavy alcohol consumption and advanced age.

Molecular Biology↗

bioRxiv2026

Rapid vaccine induction of macaque HIV-1 V2 Apex broadly neutralizing antibodies with immunogenetic signatures that are potentially translatable to humans

Due to the complex biology of HIV-1 broadly neutralizing antibodies (bnAbs), HIV-1 vaccines must target unmutated germline precursor antibodies with B cell receptors that have the appropriate specificity and genetic features to develop neutralization breadth. Previous work has shown that immunization or infection in macaques primes unmutated precursor antibodies that target the broadly neutralizing epitope composed of the second variable region and proximal glycans (termed V2-apex site) of HIV-1 envelope (Env). Macaque V2-apex bnAb precursors elicited thus far have homogeneous usage of the second reading frame of a diversity (D) gene found only in macaques, calling into question whether HIV-1 Env immunogens can elicit only this canonical, macaque-specific V2-apex antibody response, limiting human vaccination success. Here, we show that vaccination of rhesus macaques with Env conjugated to a ferritin nanoparticle (OPT4-scNP) rapidly elicited V2-apex bnAbs with distinct D gene segment usage and motifs that are not restricted only to macaques. After two priming immunizations, V2-apex antibodies with non-canonical HCDR3s exhibited greater neutralization potency and breadth than V2-apex antibodies with previously observed HCDR3 motifs. Following sequential boosting immunizations, a V2-apex bnAb lineage with a novel HCDR3-encoded EDGED motif increased its neutralization breadth, including neutralization of viral isolates bearing the N130 glycan, which shields the V2-apex from recognition. High-resolution structures of two members of this V2-apex bnAb lineage, called DH2050, defined a novel binding mode in which it used its EDGED motif to contact the C-strand peptide and the N160 glycan within the V2-apex site. Altogether, we define a novel bnAb binding mode to the V2 Apex and demonstrate that HIV-1 Env sequential vaccination elicits V2-apex bnAbs with immunogenetic signatures that are potentially translatable to the human immune system.

immunology↗

bioRxiv2026

BRIDGE-GRN: Role-Aware Bi-Tower Graph Learning with Cross-View Contrast for Directed Gene Regulatory Network Inference

Inferring directed gene regulatory networks (GRNs) from single-cell RNA sequencing (scRNA-seq) data remains difficult because expression profiles are sparse, regulatory priors are incomplete, and experimentally supported TF-target labels are limited. To address these challenges, we propose BRIDGE-GRN, a role-aware graph learning framework that separates shared graph-context encoding from directional edge decoding. BRIDGE-GRN constructs an undirected support graph from training positive regulatory evidence, learns shared gene representations with an attention-based graph encoder, and projects them into transcription factor-role and target-role embedding spaces for asymmetric TF-to-target scoring. To improve robustness under noisy and incomplete supervision, the model aligns identity and edge-perturbed graph views through cross-view contrastive regularization. We evaluated BRIDGE-GRN across mouse benchmark settings spanning five cell types, three prior-network families, and two gene-scale settings, and further examined low-supervision transfer to target domains, architectural ablations, and biological interpretability. BRIDGE-GRN achieved consistently strong performance, outperforming or matching the strongest competing baseline in most benchmark configurations. Transfer initialization improved low-shot target-domain adaptation, while ablation analyses confirmed the importance of both role-specific bi-tower projections and contrastive regularization. Biological interpretation analyses further showed role-structured embeddings, enrichment of top-ranked predictions for external regulatory support, and coherent driver-centered regulatory modules. These results support BRIDGE-GRN as a robust, transferable, and interpretable framework for directed GRN inference from single-cell transcriptomic data.

bioinformatics↗

bioRxiv2026

Comprehensive long-read transcriptome analysis uncovers alternative RNA processing feature and isoform diversity in ovarian cancer progression

Post-transcriptional processing has a crucial yet largely unresolved dynamic change and role during the malignant progression of ovarian cancer, especially due to the limited read length of short-read RNA sequencing being insufficient to capture transcript diversity. Here, we performed Iso and RNA sequencing on paired normal, primary tumor, and metastatic samples, generating a comprehensive isoform atlas of over 41,000 full-length transcripts including many unannotated isoforms. Integrative analyses revealed extensive isoform-level remodeling across disease states that often occurred without concordant alterations at the gene level, emphasizing the importance of qualitative transcript regulation. Notably, we identified isoform-level alterations with distinct biological and clinical relevance, including differential expression of the short KRAS isoform, a tumor-specific isoform switch of TMEM201, and an alternative first-exon event in FNDC3B associated with poor survival. Together, these findings provide a high-resolution map of the ovarian cancer transcriptome and illustrate how long-read sequencing exposes multiple layers of post-transcriptional and clinical insight that remain hidden in conventional expression profiling.

bioinformatics↗

bioRxiv2026

BiGAT-Fusion: Node-Wise Gated Bidirectional Graph Attention for Drug Repurposing

Drug discovery remains slow, costly, and failure-prone, motivating computational drug repurposing that prioritizes plausible drug-disease associations (DDAs). However, DDA prediction faces three stubborn challenges: (i) extreme class imbalance with few confirmed links amid vast unknowns; (ii) directional asymmetry on the bipartite drug-disease graph that standard message passing underutilizes; and (iii) static fusion of heterogeneous evidence, where fixed rules cannot adapt the relative value of feature (similarity) and topology (association) views across nodes. We present BiGAT-Fusion, a two view graph neural model that addresses these issues end-to-end. Feature view embeddings are learned on kNN graphs built from drug-drug and disease-disease similarities, while topology view embeddings are learned via a bidirectional graph attention layer that explicitly models drug[->]disease and disease[->]drug aggregation. The views are combined by type specific, node wise gates that adaptively weight evidence per node. Pair scores are produced by a residual mixture of experts head that augments a main Multi-Layer Perceptron (MLP) with a bounded low rank bilinear residual and node biases. Under repeated K-fold cross validation with validation driven selection, and with the topology graph constructed strictly from training positives, BiGAT-Fusion achieves state-of-the-art AUPRC on standard benchmarks (Gdataset, Cdataset, LRSSL, Ldataset) while remaining competitive in AUROC. Analyses of learned gates and directional attentions corroborate the models design. BiGAT-Fusion thus offers a practical, interpretable component for large-scale, computer-aided drug repurposing.

bioinformatics↗

bioRxiv2026

Gene III Ergothioneine ameliorates alcoholic fatty liver disease: A dual strategy of accelerated ethanol elimination and reducing oxidative stress

BackgroundAlcoholic fatty liver disease (AFLD) is a progressive hepatic pathology triggered by chronic ethanol consumption, serving as the initial stage of severe liver injury. Currently, there are no FDA-approved pharmacological interventions that specifically target alcohol-induced hepatic steatosis or prevent disease progression, highlighting an urgent need for effective preventive strategies. This study evaluated the preventive efficacy and underlying mechanisms of Gene III Ergothioneine (EGT) in a clinically relevant preclinical model. MethodsC57BL/6 mice were randomized into five groups: a Control group, an alcoholic fatty liver Model group, a Positive control group treated with Silybin (100 mg/kg), and three EGT treatment groups (10, 30, and 50 mg/kg). The NIAAA mouse model was utilized to induce alcoholic fatty liver. Various biochemical, histological, and molecular markers were assessed to evaluate liver damage, alcohol metabolism, lipid profiles, oxidative stress, and inflammation. ResultsGene III EGT treatment significantly ameliorated hepatic steatosis and necrosis, as confirmed by H&E and Oil Red O staining. Notably, EGT accelerated alcohol clearance, reducing serum ethanol levels by up to 54.4% in a dose-dependent manner. Furthermore, EGT restored liver function markers (ALT, AST, GGT) and corrected dyslipidemia by lowering TG, TC, and LDL-C while elevating HDL-C. Mechanistically, EGT suppressed pro-inflammatory cytokines (IL-6, IL-1 {beta}) and mitigated oxidative stress by reducing malondialdehyde (MDA) accumulation and restoring superoxide dismutase (SOD) and glutathione peroxidase (GSH-Px) activities. ConclusionGene III Ergothioneine prevents alcoholic liver injury through a dual mechanism: accelerating ethanol metabolism and enhancing hepatocyte antioxidative and anti-inflammatory defenses. These findings position EGT as a promising therapeutic candidate for AFLD management.

pharmacology and toxicology↗

bioRxiv2026

Gene III Ergothioneine Ameliorates Exercise-Induced Fatigue by Attenuating Oxidative Stress, Inflammation, and Modulating the AMPK/PGC-1α Signaling Pathway

BackgroundExercise-induced fatigue is a complex physiological phenomenon involving oxidative stress, inflammation, and metabolic disturbances. Ergothioneine (EGT), a naturally occurring amino acid with potent antioxidant properties, has garnered interest for its potential health benefits. This study aimed to evaluate the anti-fatigue effects of Gene III(R) EGT in a mouse model of exhaustive exercise and to elucidate its underlying mechanisms. MethodsMale C57BL/6 mice were randomly divided into five groups: a control group (CTL), low-dose EGT (EGT-L, 10 mg/kg), medium-dose EGT (EGT-M, 30 mg/kg), high-dose EGT (EGT-H, 50 mg/kg), and a positive control group (Coenzyme Q10, 50 mg/kg). Mice were subjected to a 4-week treadmill training protocol, followed by an exhaustive running test. We measured exercise performance and collected blood and skeletal muscle samples at multiple time points to assess biochemical markers, inflammatory cytokines, antioxidant status, and key signaling proteins. ResultsGene III(R) EGT supplementation, particularly at medium and high doses, significantly extended the time to exhaustion and running distance. Compared to the control group, EGT treatment significantly reduced post-exercise levels of lactic acid (LA), lactate dehydrogenase (LDH), and blood urea nitrogen (BUN). Furthermore, Gene III(R) EGT suppressed the exercise-induced increase in pro-inflammatory cytokines, including IL-1{beta}, IL-6, and TNF-. The anti-fatigue effect of EGT was also associated with a reduction in malondialdehyde (MDA) and an increase in the activities of superoxide dismutase (SOD) and glutathione peroxidase (GSH-Px). Mechanistically, EGT promoted the phosphorylation of AMP-activated protein kinase (AMPK) and the expression of peroxisome proliferator-activated receptor-gamma coactivator-1 alpha (PGC-1) in skeletal muscle, while also increasing the Bcl-2/Bax ratio, suggesting enhanced mitochondrial biogenesis and reduced apoptosis. ConclusionsOur findings demonstrate that Gene III(R) EGT effectively enhances exercise performance and alleviates fatigue. The underlying mechanisms involve the mitigation of oxidative stress and inflammation, as well as the activation of the AMPK/PGC-1 signaling pathway to promote mitochondrial function and cellular protection. These results highlight the potential of Gene III(R) EGT as a nutritional supplement for combating exercise-induced fatigue.

pharmacology and toxicology↗

bioRxiv2026

BOND-PEP: topology-conditioned bipartite alignment for evidence-grounded peptide binder generation

Peptide binders can modulate proteins that remain challenging for small molecules, but discovering high-affinity, selective peptides is still slow and sample-intensive. Sequence-first generators could scale design when structures are unavailable or conformationally heterogeneous, yet they often trade diversity for control: unconstrained sampling is inefficient while conditioning remains largely implicit. This limitation is exacerbated by the uneven transfer of protein language model priors to short peptides. Here we present BOND-PEP, a retrieval-augmented, bipartite-aligned, topology-conditioned framework that converts empirical binding evidence into an explicit, residue-resolved conditioning state for peptide generation. BOND-PEP shows low perplexity together with satisfactory free-generation hit rates and sequence novelty under a fair evaluation protocol and decoding budget. Compared with existing peptide generation methods, BOND-PEP achieves state-of-the-art results that match or improve upon validated peptide-protein sequence pairs. In total, BOND-PEP provides a practical, sequence-only route to controllable de novo peptide binder generation under noisy labels and distribution shift.

bioinformatics↗

bioRxiv2026

A Multimodal Single-Cell Epigenomic and 3D Genome Atlas of the Human Basal Ganglia

The basal ganglia (BG) underlie motor control, reward processing, and many neurological and psychiatric disorders, but a comprehensive epigenomic and 3D-genome atlas of the human BG is lacking. Here we present a multimodal single-cell atlas profiling DNA methylation and 3D chromatin conformation in 261,331 nuclei (snm3C-seq) across eight subregions, resolving 12 classes, 31 subclasses, and 59 groups. Harmonized under the HMBA basal-ganglia consensus taxonomy, this atlas integrates with matched RNA, ATAC-seq, and histone-modification data across five regulatory layers. We identify millions of cell-type- and region-specific differentially methylated regions enriched for distinct transcription factor motifs and link them to disease-associated heritability. Neuron-specific loops dominate cell-type-specific 3D contact remodeling, while most non-neuron-specific loops are constitutive. Among spiny projection neurons (SPNs), chromatin loops, rather than TAD boundaries, distinguish D1, D2, and eccentric SPN subclasses, with eccentric SPNs showing the most loop-level reorganization among the three. We characterize STR D2 SMYD2-HTR7 SPN, a newly recognized POU6F2 D2-SPN subtype, and reveal region-specific methylation and contact gradients of disease-associated genes, including CADM1 and PDE8B. Integrative gene-regulatory networks reconstruct cell-type-resolved enhancer-promoter links to interpret Parkinsons disease risk variants at SNCA. Finally, MERFISH spatial profiling combined with cross-species Patch-seq identifies non-SPN neuronal subtypes, including a MOXD1 striosomal STR FS PTHLH-PVALB GABA subtype with distinct electrophysiology, partitioning across the striatal matrix-striosome boundary. HighlightsO_LIA multimodal single-cell atlas maps DNA methylation and 3D genome architecture across human basal ganglia cell types and subregions. C_LIO_LINeuron-specific loops dominate cell-type-specific 3D contact remodeling in the human BG, whereas most non-neuron-specific loops are constitutive. C_LIO_LISpiny Projection Neuron (SPN) subtypes exhibit regionally organized epigenomic and 3D genome signatures that align with dorsal-ventral identities. C_LIO_LIChromatin loops are the primary distinguishing feature among D1, D2, and eccentric SPN subclasses, with eccentric SPNs being the most 3D-reorganized. C_LIO_LIIntegrated regulatory maps link cell-type-specific enhancers to disease-associated genetic risk in the human basal ganglia. C_LIO_LINon-SPN interneurons differ in distribution and in transcriptional and epigenetic identity across the matrix-striosome boundary. C_LI

neuroscience↗

bioRxiv2026

Central amygdala Isl1 neurons control biting by integrating sensory and motivational signals

Efficient orofacial motor control is essential for adaptive interactions with the environment, yet the neural substrates that regulate the force and precision of biting remain poorly understood. Here, we identify a subpopulation of neurons in the central amygdala (CeA) expressing the transcription factor Isl1 (CeAIsl1) that plays a crucial role in modulating biting behavior in mice. In vivo calcium imaging revealed that CeAIsl1 neurons are robustly activated at the onset of biting across materials of varying physical properties, with distinct neuronal ensembles selectively encoding responses to the physical properties of the objects. CeAIsl1 neuronal activity scales positively with the hardness of the object, suggesting a role in force modulation. Optogenetic activation of CeAIsl1 neurons enhances biting behavior toward edible or non-edible objects, induces fictive feeding in the absence of physical targets and exerts a reinforcing effect on behavior, whereas inhibition of CeAIsl1 neurons impaired efficient biting of solid food by reducing jaw-closing muscle activity. Projection-specific manipulations revealed that activation of CeAIsl1 projections to the parvocellular reticular formation (PCRt) and pedunculopontine tegmental nucleus (PPtg) increased the duration and frequency of biting, with CeAIsl1-to-PPTg stimulation producing a positive motivational valence. These findings uncover a previously unrecognized sensorimotor function of the central amygdala in calibrating bite force and precision, linking motivational states to skilled motor output.

neuroscience↗

bioRxiv2026

Single-cell Multiome Analysis of Chromatin State and Transcriptome in the Human Basal Ganglia

The basal ganglia play essential roles in motor control, emotion, learning and reward processing. Their dysfunction contributes to many neurological and psychiatric disorders. However, the gene regulatory programs defining basal ganglia cell-type identity and function remain poorly understood, limiting interpretation of disease-associated non-coding variants. Here, we present the first single-cell multiome atlas of histone modifications and transcriptomes across eight basal ganglia regions from neurotypical adult human donors. Joint profiling reveals cell-type-specific deployment of active and repressive cis-regulatory elements and gene regulatory networks, and suggests a combinatorial homeobox transcription factor code underlying cell identity. Integration with matched spatial transcriptomic MERFISH data uncovers regional heterogeneity of epigenomic landscapes. Comparative analysis between human and mouse medium spiny neurons uncovers conservation of core gene regulatory features. This atlas interprets non-coding risk variants of neuropsychiatric disorders and supports the development of a deep learning model to predict gene regulation and functional effects of disease-associated variants. HIGHLIGHTSO_LIJoint single-cell profiling of transcriptomes and three histone modifications across eight human basal ganglia regions characterizes active and repressive chromatin states at cell-type resolution. C_LIO_LICell-type-specific gene regulatory programs decode combinatorial homeobox TF grammar governing the identity and diversification of basal ganglia neurons. C_LIO_LIIntergrative analyses link noncoding neuropsychiatric risk variants to specific cell types, regulatory elements, and candidate target genes. C_LIO_LIA sequence-to-function deep-learning model predicts gene regulation from DNA sequence and prioritizes functional disease-associated variants. C_LI

genomics↗

bioRxiv2026

Single-Cell Atlas of Transcription and Chromatin States Reveals Regulatory Programs in the Human Brain

Directly measuring chromatin states alongside transcription is essential for understanding how cell-type-specific regulatory programs are established and maintained in the adult human brain. We present a large-scale single-cell multimodal atlas generated by jointly profiling transcriptome with active (H3K27ac) and repressive (H3K27me3) histone modifications across 18 brain regions. We profile >750,000 nuclei spanning 160 cell types and integrate these data with chromatin accessibility, DNA methylation, 3D genome architecture, and spatial transcriptome. This framework annotates >500,000 regulatory elements and resolves cell-type-specific chromatin states. We link enhancers to target genes, infer gene regulatory networks, and classify chromatin interactions, revealing neuron-enriched long-range Polycomb repression of developmental genes. Integrating these maps with GWAS data and sequence-based model prioritizes noncoding variants, effector genes, and vulnerable cell types for neuropsychiatric disorders. Finally, cross-species comparisons show conserved activation but more divergent repression. Together, this study provides a functional reference for interpreting noncoding variants, epigenetic memory, and brain organization. HIGHLIGHTSO_LIJoint single-cell profiling of transcriptomes with active or repressive histone modification in >750,000 nuclei across adult human brain. C_LIO_LIChromatin state annotation of >500,000 candidate cis-regulatory elements distinguishes active enhancers from accessible and Polycomb-repressed regions. C_LIO_LICell-type-resolved regulatory networks and sequence-based deep learning model prioritize functional neuropsychiatric risk variants. C_LIO_LISpatial epigenomic imputation reveals laminar layer-specific Polycomb repression programs. C_LIO_LIIntegration with 3D genome architecture reveals neuron-specific super long-range chromatin loops silencing early developmental genes. C_LIO_LIEvolutionary analysis uncovers conserved active regulatory grammar but divergent repressive landscape. C_LI

genomics↗