Biology subjects

Li, K.

Publications and source records attributed to Li, K..

At least 73 records · Page 4Linked to original sources

bioRxiv2026

Structure-guided discovery and engineering of miniature CRISPR-Cas12m for epigenome editing

CRISPR-based epigenome editing represents a programmable strategy to precisely modulate gene expression, holding immense promise for therapeutic applications. However, the large size of the dCas proteins substantially impedes the delivery via adeno-associated virus (AAV) vectors. Here, through iterative bioinformatics analysis, structure-guided predictions, and functional assays, we identified and characterized PmCas12m, a novel miniature subtype V-M CRISPR-Cas12m. PmCas12m exhibited flexible 5-YTN-3 PAM-dependent recognition and robust double-stranded DNA binding properties, while lacking DNA cleavage activity, thus positioning it as an ideal tool for epigenome editing. Cryogenic electron microscopy (cryo-EM) structures of PmCas12m unveiled its unique molecular mechanism of DNA binding facilitating interference. Guided by these structural insights, we employed deep mutational scanning (DMS) and protein engineering to develop xCas12m, a hypercompact variant with highly potent and specific epigenome editing capabilities in human cells. We further constructed the xCas12m-CRISPRoff platform in a single AAV vector, which achieved durable epigenetic silencing and effective inhibition of hepatitis B virus (HBV) infection in a mouse model. Collectively, these findings establish xCas12m as a versatile epigenome editing platform with transformative potential for treating diseases, paving the way for clinical translation of epigenetic therapies.

bioengineering↗

bioRxiv2026

Uncovering zebrafish embryonic proteome dynamics across 16 time points during the first 24 hours of development

Defining how proteins change over developmental time is amenable to studies deciphering regulatory genetic networks in vertebrate development, biology, and pharmacology. In an approach toward such quantitative studies of dynamic network behavior, we produced an atlas using the mass spectrometry-based method to investigate protein expression changes across 16 time points from the zygote to the early pharyngula stage zebrafish embryos. We systematically summarize 8 clusters for interrogating changes in protein expression associated with the development of zebrafish embryos. Specifically, we identified a class of zinc finger-related transcription factors primarily located on the long arm of chromosome 4, which are highly expressed during zygotic genome activation. Furthermore, we highlight the power of this analysis to assign developmental stage-specific expression information to chromosomes and tissues. Time-resolved analyses reveal significant discordance between differential transcript and protein expression, whereas no time lag is observed for proteins involved in stable and fundamental biological processes, such as metabolism (e.g., Ppt2a and Gatm), cytoskeletal organization (e.g., Col18a1), and the translation machinery (e.g., Eif4enif1). This atlas offers high-resolution and in-depth molecular insights into zebrafish development, providing a resource for developmental biologists to generate hypotheses for functional analysis of proteins during early vertebrate embryogenesis. HighlightsO_LIA global protein expression database with high time resolution is created for zebrafish embryos. C_LIO_LIDistinct patterns of protein expression correlate with biological processes. C_LIO_LITranscription factors have a burst of expression from the gastrulation stage. C_LIO_LIDevelopmental stage-specific protein expressions were assigned to chromosomes and tissues. C_LIO_LIHigh-resolution embryonic transcriptome and proteome datasets were compared and connected. C_LI

developmental biology↗

bioRxiv2026

HitAnno: Atlas-level cell type annotation based on scATAC-seq data via a hierarchical language model

The single-cell assay for transposase-accessible chromatin using sequencing (scATAC-seq) has emerged as a core technology for dissecting cellular epigenomic heterogeneity and gene regulatory programs. With the emergence of atlas-level scATAC-seq datasets, cell type annotation increasingly faces challenges arising from unprecedented data scale and increased cell-type diversity, which together place stringent demands on model reliability and robustness. Here, we introduce HitAnno, a hierarchical language model capable of accurate and scalable cell type annotation in atlas-level scATAC-seq data. Leveraging selected cell-type-specific peaks to construct "cell sentences", HitAnno employs a two-level attention mechanism that captures accessibility profiles hierarchically. Extensive evaluations show that HitAnno robustly annotates both major and rare cell types across multiple settings, including intra-dataset, cross-donor and inter-dataset annotation. The hierarchical attention mechanisms of the model reveal co-accessibility patterns among peaks and dependencies across higher-order peak sets, ensuring an interpretable annotation process. Training on a 31-cell-type human atlas, HitAnno can directly annotate new query datasets without retraining and is accessible through an online interface. Our model identifies heterogeneous subgroups within mixed labeled cells from unseen datasets, demonstrating its potential to assist researchers in refining existing cell atlases.

bioinformatics↗

bioRxiv2026

Enteric sensory neurons for nutrient detection and gut motility

The enteric nervous system (ENS) orchestrates gastrointestinal reflexes and brain-gut communication via molecularly diverse neurons. Among these, intrinsic primary afferent neurons (IPANs) are essential for detecting luminal nutrients and irritants, yet their molecular identities, sensory properties, and functions remain poorly resolved. Here, we establish a segment-resolved single-cell atlas of the murine ENS, including a comprehensive characterization of the gastric ENS. This resource defines a refined taxonomy of enteric neurons and glia and enabled the development of a genetic toolkit for molecularly defined IPANs. Using chemogenetics and calcium imaging, we discovered that myenteric neurons detect a wide range of nutrients, irritants, and cytokines. Nutrient detection depends on a functional connection between chemosensory epithelial cells and enteric neurons mediated by 5-HT--HTR3 axis. Through optogenetic analysis, we demonstrated segment-specific regulation of gut motility by different IPANs. Our work establishes a genetic and physiological framework for enteric-specific sensory mechanisms.

neuroscience↗

bioRxiv2026

MetaSAG: A Tool for Multi-level Exploration and Taxonomic Analysis of Microbial Single-Amplified Genomes

Microbial single-amplified genome (SAG) sequencing technologies have elevated microbial research resolution to the single-cell level. However, neither upstream data processing nor downstream analysis has been fully developed, greatly limiting the research in strain level. Herein, we developed MetaSAG (Multi-level Exploration and Taxonomic Analysis of microbial Single-Amplified Genomes), which enables accurate and rapid taxonomic classification of microbial SAGs. MetaSAG outperforms existing method in species classification certainty, computational efficiency, and sensitivity of low abundance species identification. In addition, MetaSAG enables species-level functional analysis, as well as strain-level evolutionary analysis. With the help of MetaSAG, we discovered the parasitic relationship between phages and bacteria, identifying multiple susceptible bacteria and a broad spectrum of phages. Furthermore, we developed MetaK-Lytic (k-mer-based meta-learning framework to predict phage lytic ability) to achieve accurate prediction of phage lytic activity based on 31-mer short sequences, which is well adapted to the characteristics of incomplete SAG sequences. Overall, we offer a comprehensive integrated tool that can parse microbial SAG data from raw data to the strain level to decipher the functional ecology of microbial dark matter, with broad implications for microbial ecology and phage therapy (https://github.com/liangcheng-hrbmu/MetaSAG). Significance StatementThis work provides a comprehensive framework for high-resolution SAG data analysis. The developed pipeline improves taxonomic annotation sensitivity and speed. Strain-level tracking enables dynamic evolutionary and functional insights, while single-cell bacterial-virus network reconstruction reveals precise interaction patterns. The novel annotation-free short sequence-based MetaK-Lytic facilitates functional prediction of uncharacterized phage sequences. Integrated into the MetaSAG platform, these tools deliver a streamlined, multi-level solution for interpreting SAG data, advancing studies in microbial ecology, evolution, and virology.

microbiology↗

bioRxiv2026

Circular RNA identification using a genomic language model and a small number of authenticated examples

Genomic language models (gLMs) hold great promise for deciphering biological sequences, yet their effectiveness is hindered by the limited number of experimentally verified examples available for model training, a ubiquitous bottleneck for supervised machine learning. To overcome this challenge, we developed circFormer, the first gLM-driven approach for circular RNA (circRNA) identification. circFormer integrates curriculum learning with gLM fine-tuning: a Nucleotide Transformer model is first trained on a small set of validated circRNAs, the resulting model is used as a teacher to score [~]2.3 million noisy candidates, and the model is then fine-tuned with the noisy candidates along with their scores to improve prediction. Operating either as a standalone predictor or as a filter for existing pipelines, circFormer consistently outperformed traditional machine-learning approaches in accuracy and robustness. Among 50 circFormer-selected candidates that were overlooked by most existing tools, experimental validation using RNase R digestion and RT-qPCR confirmed 94.1% (32/34) of the evaluable candidates as genuine circRNAs. To enhance interpretability, we introduced a model-agnostic, dual-level explainable AI strategy that reveals mechanistic signatures of circRNA formation. circFormer provides a scalable, interpretable, and generalizable framework for converting noisy high-throughput data into reliable functional annotations, highlighting a practical path forward for gLM-based genomics in data-scarce settings.

bioinformatics↗

bioRxiv2026

Chd8 haploinsufficiency leads to molecular layer heterotopias and age-dependent cortical expansion

Mutations in the chromatin remodeler CHD8 are associated with autism and macrocephaly. While mouse models of Chd8 haploinsufficiency recapitulate brain overgrowth, the specific cellular mechanisms and developmental timing that lead to these anatomical abnormalities remain poorly understood. Here, we conducted 3D imaging of Chd8V986*/+ mouse brains using magnetic resonance imaging followed by tissue clearing and cellular resolution light-sheet microscopy across embryonic and postnatal developmental stages. We found that brain overgrowth occurs postnatally, driven by an increase in non-neuronal cells prior to volumetric expansion. Unexpectedly, we identified prevalent molecular layer heterotopias (MLH) within the frontal cortex of Chd8V986*/+ mice composed of neurons breaking through the pial surface during embryonic development and persisting throughout life. Increased incidence of MLH, previously identified in individuals with idiopathic autism, was replicated across three independent Chd8+/- mouse models and present across multiple genomic backgrounds, establishing aberrant neuronal migration as a core feature of Chd8 haploinsufficiency.

neuroscience↗

bioRxiv2026

A foundation AI model enhances electron microscopy image analysis

Electron microscopy (EM) and advanced volume EM have extensive applications in deciphering cellular ultrastructures for life sciences. However, quality issues of EM images significantly impede the precise analysis and discovery of nanoscale biological structures. Here, we present an unsupervised Foundation Model for Five Tasks (DF5T) model for image enhancement through denoising, deblurring, super-resolution, two-dimensional (2D) inpainting, and three-dimensional (3D) isotropic restoration. DF5T was trained on an extensive dataset of multi-source membrane-bound organelle EM images, consisting of a total of over 2.25 million images. On unseen data, DF5T outperformed the existing state-of-the-art models in all five tasks. Furthermore, DF5T substantially compensates for missing three-dimensional structural information and reveals significant ultrastructural alterations in organelle geometry following chemical treatment. We demonstrate that DF5T significantly enhances electron microscopy images quality by improving the accuracy of downstream organelle segmentation and 3D structural restoration, thereby contributing to future advances in biological research.

cell biology↗

bioRxiv2026

Predicting recovery trajectories and injury severity following partial crush spinal cord injury in mice

The partial crush spinal cord injury (SCI) model enables preclinical testing of experimental therapies in mice, but substantial inter-animal variability in recovery outcomes confounds efficacy assessments. Here, we used open field behavioral data collected during the first 3 days post partial thoracic SCI to generate an Acute Functional Score (AFS) that defined three subgroups with divergent recovery trajectories. Applying latent class growth analysis and growth mixture modeling to open field and grid walk testing data, we demonstrated 83-92% prediction accuracy for AFS-defined recovery trajectories. The three subgroups differed significantly in treadmill kinematics and histological assessments of lesion size and astrocyte bridging. Applying the recovery trajectory framework to mice receiving saline or biomaterial vehicle injections at 3 days post-SCI revealed robust predictive accuracy while exposing disproportionate injury severity distributions between experimental groups. The approach enables individualized post-SCI recovery characterization that can neutralize procedural bias, minimize animal numbers, and provide a probabilistic basis for evaluating whether interventions enhance or suppress wound repair processes. Our findings establish a foundation for improving preclinical SCI study design and accelerating identification of effective therapies.

neuroscience↗

bioRxiv2026

A vagal gut-brain axis for feeding induced gastric secretion

Feeding begins with food-seeking and ends with nutrient absorption and waste elimination. Although meal-related sensory cues are known to shape feeding behavior, how they engage neural circuits to control digestion remains unclear. We dissected vagal gut-brain-gut pathways that mediate meal-induced gastric secretion. Vagal afferent neurons innervating the stomach were necessary and sufficient for secretion. Among those, Glp1r and Sst neurons robustly drove acid, pepsin, and gastrin secretion, whereas Oxtr and Vip neurons had little effect. Piezo2 was expressed in Glp1r mechanosensory afferents, and conditional deletion of Piezo channels attenuated distension-evoked secretion. On the efferent limb, genetically defined vagal motor neuron subsets differentially recruited secretory outputs: Cck neurons broadly increased acid, pepsin, and gastrin, whereas Grp neurons preferentially drove acid secretion. Finally, enteric circuit dissection revealed modular output control: Calb2 enteric neurons potently drove secretion with minimal motility effects, Grp enteric neurons coupled secretion to motility, and Cysltr2 enteric neurons selectively modulated acid output. Together, these results define a cell-type resolved vagal-enteric architecture for gastric control and show that secretory and motor programs can be segregated across dedicated enteric neuron subtypes.

neuroscience↗

bioRxiv2026

Minimal biophysical rules are sufficient for the emergence of computational intelligence at the neuronal scale

How intelligence emerges from the brains complex microscopic physical system is a central question for neuroscience and artificial intelligence. Constrained by the genomic bottleneck that precludes synapse-by-synapse specification, we propose and validate a microscopic structure-function concise-constraint sufficiency hypothesis. We develop the Neuro-Informed Generative Connectome (NIGC) framework, and show that connectomes generated under a concise set of biophysical constraints (geometric embedding, node propensity modulation, a global energy budget and maximum-entropy selection) closely match the structural statistics of a measured mouse V1 microcircuit (similarities, 0.997). In parallel, using the generated connectome as the fixed reservoir of an echo state network (ESN), training only a linear readout achieves 90% accuracy on an auditory multi-classification task. Moreover, multidimensional biologically consistent functional phenotypes, such as hierarchical transmission delays and low-dimensional spatiotemporal trajectories, are obtained without fitting functional matrices or time courses. Further, by combining single-constraint ablations, pathological perturbations and cross-modal validation, we clarify how specific structural constraints map onto functional consequences. Together, these results delineate sufficient conditions for computational intelligence emergence at the microscopic scale, and provide an auditable benchmark for first-principles understanding of brain construction.

neuroscience↗

bioRxiv2026

Fully Phased Telomere-to-Telomere Assemblies for Thoroughbred Horse and Donkey Haplotypes derived from a Mule Illuminate the Peculiar Evolution of Equid Centromeres

We present telomere-to-telomere genome assemblies of a Thoroughbred horse and a donkey derived from their mule offspring. Now adopted and annotated by NCBI as reference genomes, these assemblies resolve previously inaccessible regions, including satellite arrays, duplications, and telomeres. Equids are known to exhibit an uncoupling between satellite DNA and centromeric function. The completeness of these assemblies enabled annotation of both satellite-based and satellite-free centromeres, as well as non-centromeric satellite loci, revealing notable centromeric plasticity. They also allowed detailed characterization of the variable binding domains of CENP-A--the epigenetic determinant of centromere identity--and CENP-B, whose association with CENP-A, previously considered typical based on a few model organisms, is absent in equids. Comparative analyses of satellite repeats and centromere positions provide new insights into the accelerated karyotypic reshuffling in equid evolution. These assemblies represent foundational resources for equid genomics and support ongoing initiatives such as the Equine Pangenome Project.

molecular biology↗

bioRxiv2026

Comprehensive top-down mass spectral repository enables pan-dataset analysis and top-down spectral prediction

Mass spectral libraries have become essential resources for training deep learning (DL) models for spectral prediction and de novo sequencing in bottom-up mass spectrometry (BU-MS). Compared with BU-MS, top-down MS (TD-MS) offers unique advantages for characterizing intact proteoforms by analyzing proteoforms without enzymatic digestion. Despite these advantages, large-scale spectral libraries for TD-MS are currently lacking. Here we present TopRepo, the first comprehensive and publicly available repository of TD-MS spectra, comprising more than 12 million spectra acquired from 12 species across seven types of mass spectrometers. Using TopRepo, we constructed a large-scale top-down spectral library containing over 5 million spectra with proteoform and fragment-ion annotations. We demonstrate that TopRepo enables pan-dataset analyses of N-terminal processing, mass shifts, and other proteoform characteristics identified by TD-MS. Furthermore, we show that the TopRepo spectral library substantially improves proteoform identification through spectral library searching and supports the training of DL models for high-accuracy top-down spectral prediction.

bioinformatics↗

bioRxiv2026

A specialized ARGONAUTE enables trans-species RNA interference in plant immunity

Trans-species RNA interference (tsRNAi), in which plants produce small RNAs (sRNAs) to silence target genes in pathogens, has emerged as a promising strategy for disease control. However, whether tsRNAi constitutes an endogenous, regulated immune response remains unclear. Here, we show that ARGONAUTE10 (AGO10) plays a critical role in pathogen-induced tsRNAi. Loss of AGO10 in Arabidopsis abolished pathogen gene silencing during infection, leading to hypersusceptibility to oomycete and fungal pathogens. Importantly, AGO10 rapidly responds to pathogen infection through increased protein accumulation and re-location into discrete cytoplasmic condensates, thus promoting the production of trans-species sRNAs at the pathogen infection sites. This immune responsiveness relies on the N terminal intrinsically disordered region (IDR) of AGO10, which is responsible for sensing and responding to immune activation. Specific features in the IDR partitions AGO10 into two deeply diverged subgroups, AGO10a and AGO10b, with the immune responsiveness and defense function evolutionarily conserved in AGO10a but not AGO10b. Together, these findings establish tsRNAi as a bona fide, evolutionarily conserved immune response and position AGO10 as a signal-responsive hub linking pathogen perception to tsRNAi-based defense.

plant biology↗

bioRxiv2026

Epigenetic activation of CHST2 by promoter hypomethylation promotes progression of triple-negative breast cancer

Epigenetic dysregulation is a hallmark of triple-negative breast cancer (TNBC), yet the oncogenic relevance of specific epigenetically activated genes remains poorly defined. In this study, we demonstrate that carbohydrate sulfotransferase 2 (CHST2) is aberrantly upregulated in TNBC across multiple transcriptomic analyses as a consequence of promoter hypomethylation. Elevated CHST2 expression is associated with aggressive clinicopathological characteristics and has been linked to unfavorable patient outcomes in independent cohorts. Functional analyses reveal that CHST2 primarily promotes TNBC cell migration and invasion, while genetic silencing of CHST2 markedly attenuates these malignant phenotypes. Importantly, mutational disruption of CHST2 catalytic activity abolishes its pro-migratory effects, indicating that sulfotransferase activity is essential for CHST2-driven invasiveness. Mechanistically, promoter hypomethylation is experimentally supported as a driver of CHST2 transcriptional activation, linking epigenetic deregulation to enhanced tumor aggressiveness. In addition, tumors with high CHST2 expression exhibit distinct immune-related transcriptional features within the tumor microenvironment. Collectively, these findings identify CHST2 as an epigenetically activated driver of invasive behavior in TNBC and highlight its potential value as a biomarker and therapeutic target for aggressive disease.

cancer biology↗

bioRxiv2026

AI-driven discovery and engineering of human endogenous nanocage proteins for mRNA delivery

Safe and effective gene delivery remains a central challenge for therapeutic applications. While non-viral and viral vectors have enabled substantial progress, their reliance on non-human components often triggers immune responses, limiting their use in chronic treatments. Here, we developed DeepDelivery, an artificial intelligence-driven platform to repurpose human proteins for mRNA delivery. An unbiased screening of the human proteome nominated 512 candidates, with experimental validation confirming that 80% of top-ranked hits form mRNA-encapsulating particles and mediate efficient functional delivery in human cells without provoking detectable inflammation. Notably, multiple tripartite motif (TRIM) family proteins, typically linked to antiviral responses, exhibited strong assembly and delivery activity. Quantitative analysis and interpretation of the model revealed structural domains that govern nanocage formation, enabling domain-guided engineering of TRIM25 variants with enhanced function. Our work establishes a generalizable framework for discovering human-derived delivery vehicles and provides a path toward programmable, non-immunogenic mRNA therapeutics.

synthetic biology↗

bioRxiv2026

Looking backward for the future: long-term population recovery, habitat associations, and future climatic vulnerability of the critically endangered Sino-Mongolian beaver (Castor fiber birulai) in China

Despite the successful population recovery of the Eurasian beaver (Castor fiber) across much of Eurasia, its subspecies, the Sino-Mongolian beaver (C. f. birulai), remains critically endangered, with an estimated population of approximately 1,500 individuals confined to a small number of isolated and fragmented refugia along the China-Mongolia border. Effective conservation of this highly threatened subspecies requires a holistic perspective that integrates constraints on population dynamics, habitat associations, and future climatic vulnerability. Here, we combined systematic annual field surveys conducted between 2003 and 2023 with historical survey records from 1975 to 1989 in northern Xinjiang, China, to synthesize long-term spatiotemporal population dynamics, evaluate habitat preferences based on nine local environmental variables, and assess future climatic vulnerability using ensemble species distribution models (SDMs) under projected climate change scenarios. We detected a significant and phased population recovery, with beaver colony numbers increasing from 27 (approximately 100 individuals) in 1975 to 227 (681-908 individuals) in 2023. This recovery closely corresponded with major conservation milestones, including the establishment and upgrading of nature reserves, strengthened legislative protection, and enhanced multi-stakeholder collaboration. Habitat analyses further indicated that the Sino-Mongolian beaver preferentially occupied areas characterized by minimal anthropogenic disturbance and stable hydro-geomorphic conditions. Critically, SDM projections revealed that only 14% of the current study area presently exhibits high climatic suitability, and these highly suitable habitats are expected to disappear entirely by the 2050s. Together, our findings provide a comprehensive overview of the historical population recovery and conservation trajectory of the Sino-Mongolian beaver in China, and offer robust scientific support for developing adaptive management strategies in the face of ongoing climate change and increasing human pressures.

ecology↗

bioRxiv2026

BOLE: A Knowledge-Enhanced Multi-Agent Framework for Intelligent Genomic Breeding

Genomic breeding has become increasingly data-intensive, yet the practical integration of heterogeneous bioinformatics tools into coherent analytical workflows remains a major bottleneck. To address this, we present BOLE, a knowledge-enhanced multi-agent AI framework for autonomous genomic breeding analysis. By integrating a structured script knowledge base with constrained large language model reasoning, BOLE translates high-level analytical intent into validated and executable workflows. Specialized agents collaboratively perform intent interpretation, task decomposition, dependency reasoning, workflow synthesis, and controlled code generation to assemble validated, executable workflows without predefined pipelines. BOLE supports core genomic breeding tasks, including genome-wide association studies, heritability estimation, germplasm evaluation, and genomic selection while preserving artifact-level provenance to ensure reproducibility. Implemented as a web service, BOLE enables end-to-end genomic breeding analysis from raw data to actionable results for non-expert users. This work establishes a generalizable paradigm in which knowledge-driven multi-agent AI bridges the gap between mature analytical methods and their real-world application in genomic breeding.

genomics↗