Biology subjects

Liu, Y.

Publications and source records attributed to Liu, Y..

At least 235 records · Page 13Linked to original sources

bioRxiv2026

C1qa⁺ muscularis macrophages maintain enteric synaptic homeostasis to regulate gastrointestinal motility

The enteric nervous system (ENS) is a complex peripheral neural network that coordinates gastrointestinal motility through highly organized synaptic communication. Although tissue-resident muscularis macrophages (MMs) closely associate with enteric neurons, whether they regulate enteric synaptic organization remains unknown. In the central nervous system (CNS), microglia sculpt neural circuits through complement-dependent synaptic remodeling, raising the possibility that analogous neuroimmune mechanisms operate in the gut. Here, we identify a previously unrecognized role for C1qa{square} MMs in regulating enteric synaptic homeostasis and gastrointestinal motility. Using macrophage-specific constitutive and inducible C1qa deletion models, single-cell RNA sequencing, enteric synaptosome proteomics, physiology, and advanced imaging, we demonstrate that loss of MMs-derived C1qa increases enteric synaptic density without altering neuronal numbers. C1qa deficiency induced broad transcriptional changes in enteric neurons and macrophages, including altered synapse-associated, lysosomal, and endocytic programs. Proteomic analysis revealed that enteric synapses share a conserved molecular architecture with brain synapses while exhibiting distinct gastrointestinal-specific complement-associated synaptic networks enriched for structural and receptor-localization pathways. Functionally, macrophage-specific C1qa deletion altered excitatory and inhibitory enteric neurotransmission, enhanced cholinergic signaling, reduced nitrergic responses, and accelerated gastrointestinal transit, while smooth muscle responsiveness remained preserved. C1qa{square} MMs displayed transcriptional and functional features consistent with a phagocytic synapse-remodeling phenotype, including enrichment of complement, lysosomal, and engulfment pathways. Loss of C1qa impaired macrophage phagocytic activity both in vitro and in vivo and was associated with synapse accumulation and altered macrophage morphology. Importantly, inducible deletion of C1qa in adulthood recapitulated the synaptic and motility phenotypes, demonstrating that C1qa{square} MMs continuously regulate enteric synaptic organization beyond development. Together, these findings identify a complement-dependent neuroimmune mechanism that regulates enteric circuit organization and gut motility, establishing MMs as active modulators of adult ENS synaptic homeostasis. O_FIG O_LINKSMALLFIG WIDTH=200 HEIGHT=117 SRC="FIGDIR/small/729640v1_ufig1.gif" ALT="Figure 1"> View larger version (52K): org.highwire.dtl.DTLVardef@16aadd8org.highwire.dtl.DTLVardef@bb74f3org.highwire.dtl.DTLVardef@fb040corg.highwire.dtl.DTLVardef@1c6e4a5_HPS_FORMAT_FIGEXP M_FIG C_FIG

neuroscience↗

bioRxiv2026

Telomeric heterochromatin acts in trans to promote meiotic centromere assembly via Aurora B kinase recruitment

During meiosis, chromosomes face a paradox: the machinery that ensures reductional chromosome segregation also destabilizes centromeres by dismantling kinetochores, risking chromosome missegregation. Here we show how cells resolve this crisis through an unexpected activity of the telomere bouquet. We demonstrate that the bouquet transfers heterochromatin components to pericentromeres, which in turn recruit the Aurora B kinase to direct centromere reassembly. The heterochromatin protein Swi6HP1 relocates from telomeres to centromeres, enabling Haspin kinase-dependent phosphorylation of histone H3 and consequent enrichment of the chromosomal passenger complex, which includes the Aurora B kinase. Aurora B then phosphorylates core centromere proteins, including CenpA and CenpC, to promote kinetochore reassembly. Phosphomimetic mutants of CenpA or CenpC bypass the telomere-heterochromatin-Haspin pathway, demonstrating that Aurora B-mediated phosphorylation is sufficient for reassembly. This function is conserved in mitotically proliferating cells subjected to centromere dismantlement. Our findings establish a safeguarded system that couples meiotic nuclear architecture to centromere identity and reveal a fundamental role for the Aurora B kinase in centromere assembly, beyond its canonical function in correcting kinetochore-spindle attachment errors.

cell biology↗

bioRxiv2026

Long-read RNA-seq resolves isoform-level and context-specific regulatory architecture of complex traits in cattle

Traditional short-read RNA-seq lacks the resolution to fully capture full-length isoforms and complex alternative splicing, leaving a "missing regulation" gap in existing molecular atlases. Here, we present a population-scale, multi-omics dissection of bovine complex traits using matched deep whole-genome sequencing (30x), long-read (ONT) and short-read RNA-seq, as well as metabolome data from 432 dairy cows across four distinct lactation stages. We expanded the bovine transcript atlas with 13,177 novel isoforms and fine-mapped 1,088,900 regulatory effects across 11 diverse molecular phenotypes. Notably, long-read mapping uncovered 606 genes with isoform expression quantitative trait loci (12.8%, eQTL) and 1,116 genes with detailed splicing QTL (sQTL) (31.8%) that were missed by conventional short-read mapping. We further showed that lactation-specific regulation was profoundly mediated by the metabolic microenvironment (46.8%) and cellular composition (4.3%). Across genome-wide associations (GWAS) of 90 bovine complex traits, isoform-level and context-specific regulatory effects captured GWAS loci missed by short-read eQTL and sQTL analyses, increasing colocalized loci from 35.2% to 60.3%. These regulatory variants also exhibited evolutionary constraints with human immune and metabolic loci. Our study establishes an invaluable resource (https://cattleblr.farmgtex.org) demonstrating that long-read transcriptomics is essential for resolving the complexity of regulatory mechanisms underlying mammalian physiology.

genetics↗

bioRxiv2026

Discovery of ARQ-501 as a potent PafA inhibitor that elevates KatG levels to potentiate isoniazid in mycobacteria

The prokaryotic ubiquitin-like protein (Pup) conjugation system (PPS), which is essential for Mycobacterium tuberculosis (Mtb) virulence but absent in humans, presents an attractive drug target. Here, we report the discovery of ARQ-501, a quinone-based, covalent, substrate-competitive inhibitor of the Pup ligase PafA. ARQ-501 exhibited potent anti-mycobacterial activity against Mtb under host-mimicking stress conditions and within macrophages. We further identified the catalase-peroxidase KatG, essential for activation of the frontline prodrug isoniazid (INH), as a pupylation substrate. ARQ-501 inhibits KatG pupylation, causing its accumulation and creating a selective synergy with INH. This quantity over quality mechanism successfully rescued INH activation by the clinically prevalent KatG S315T mutant in enzymatic assays and enhanced INH efficacy against clinical S315T isolates to variable degrees. This work identifies a novel class of PafA inhibitors and a previously unrecognized role of pupylation in regulating KatG, offering a potential therapeutic avenue to combat drug-resistant tuberculosis.

microbiology↗

bioRxiv2026

The Impact of Smoking Status on the Genomic Landscape of Lung Squamous Cell Carcinoma

PurposeComprehensive genomic profiling (CGP) has changed the treatment paradigm for non-small cell lung cancer (NSCLC) with the advent of molecularly targeted therapies for actionable genomic alterations (AGA). Despite this, the use of CGP is suboptimal, particularly in squamous cell lung cancer (sqNSCLC), which is more closely associated with smoking exposure and a lack of AGAs. We hypothesized that the prevalence of AGAs is inversely correlated with the chronicity and extent of smoking exposure in patients with sqNSCLC. Experimental DesignWe retrospectively evaluated all patients with liquid biopsy testing via Guardant 360CDX or Guardant360 in the context of any sqNSCLC diagnosis at the City of Hope Comprehensive Cancer Center between 10/2020 and 7/2023. The data was obtained on 2/23/24. Social and clinical histories were evaluated to assess the frequency of AGAs in patients with no or remote smoking history. ResultsOf the 56 patients in the initial evaluation, 24% (n=13) were non-smokers or remote smokers (greater than 20 years from cessation). Of these 13 patients, eight (61.5%) harbored AGA. Of these 8 patients, alterations observed included EGFR exon 19 deletion (50%, n=4), MET exon 14 skipping mutation (25%, n=2), EGFR G719S (13%, n=1), EGFR E114K (13%, n=1). Of those patients harboring AGAs that received NCCN-concordant matched targeted therapy, the objective response rate (ORR) with targeted agents was 50% and the clinical benefit rate (CBR) was 83.3%. ConclusionsThese data support the use of CGP in sqNSCLC particularly in patients with remote or no smoking exposure. Statement of translational relevanceThese data demonstrate high frequency of actionable genomic alterations (AGAs) in patients diagnosed with squamous cell lung cancer (sqNSCLC) with remote or no smoking history. Specifically, enrichment of EGFR and MET gene alterations were observed. These findings support the use of comprehensive molecular profiling in sqNSCLC. Furthermore, treatment outcomes demonstrate frequent objective responses and high clinical benefit rate supporting the use of targeted therapies in sqNSCLC harboring AGAs. This analysis provided rationale for further research of larger datasets investigating therapeutic approaches in sqNSCLC, which may have significant implications for consensus guideline recommendations and routine clinical practice.

cancer biology↗

bioRxiv2026

A Human Vocal Fold Organ-On-Chip for Studying Platform-Dependent Mucosal Responses to Particulate Matter

BackgroundCoarse particulate matter (PM10) deposits at the vocal fold (VF) mucosa, yet upper airway responses remain poorly characterized. Existing in vitro VF models use monocultures that lack the stratified epithelium, lamina propria, and physiological perfusion. We developed a chip-based co-culture model of human VF mucosa and applied it to acute PM10 exposure. MethodsVocal fold organ-on-chip (VF-OOC) paired primary VF fibroblasts with either immortalized laryngeal epithelial cells (iLEC) or induced pluripotent stem cells (iPSC)-derived VF epithelial cells. Transwell and 2D chip cultures were controls. Epithelia matured at an air-liquid interface on fibroblast-embedded collagen over a perfused microchannel. PM10 urban dust (0 to 400 {micro}g/mL) was applied for 24 hours. Responses were assessed by histology, immunofluorescence, transmission electron microscopy, qPCR, and ELISA. ResultsVF-OOC produced a thicker stratified epithelium with upregulated barrier, mucin, and extracellular matrix genes versus transwell controls. Intercellular junctions and basement membrane matched adult human VF mucosa. PM10 remained at the epithelial surface across all doses. TranswelliLEC downregulated basal markers (TP63, KRT5, KRT14). VF-OOCiLEC upregulated MUC1 and HAS3, consistent with an adaptive mucosal response. VF-OOCiPSC additionally induced suprabasal, junctional, extracellular matrix, and cytokine genes largely absent in iLEC and transwell formats. ConclusionsVF-OOC reproduced key features of native human VF mucosa and captured distinct, platform-specific responses to PM10 that differed by epithelial cell source. By introducing new approach methodologies (NAMs) into laryngology, this platform extends respiratory toxicology to the upper airway beyond bronchial and alveolar compartments and allows mechanistic studies of exposure-linked diseases such as laryngitis.

cell biology↗

bioRxiv2026

Assessing and Optimizing Low-Frequency Somatic Mutation Detection: A Multi-Platform High-Throughput Sequencing Perspective

The availability of multiple commercial short-read sequencing platforms necessitates systematic cross-platform performance comparisons, particularly for challenging applications such as low-frequency somatic mutation detection. Here, a large-scale targeted sequencing dataset from five Genome in a Bottle (GIAB) human genomic DNA reference standards, HG001 to HG005, alongside Twist Biosciences cfDNA reference standards featuring 1% variant allele frequency (VAF), was generated by six platforms (NovaSeq 6000, NovaSeq X, FASTASeq 300, GenoLab M, SURFSeq 5000, and MGISEQ-T7). To build a realistic benchmark while keeping authentic sequencing backgrounds, we developed PosMix, a simulating tool that generates position-specific VAFs. To overcome the limitations of conventional variant callers (high recall with poor precision for VarScan2, higher precision with lower recall for Strelka2/Mutect2), we developed SomaticXGB, a machine learning-based caller. In this study, SURFSeq 5000 consistently exhibited the lowest error rates and achieved superior accuracy for VAFs as low as 0.5%, outperforming all other sequencing platforms. On the other hand, SomaticXGB attained F1 scores of approximately 0.92 on simulated datasets with VAFs ranging from 0.5% to 1.5% and 0.89 on Twist 1% standards, substantially outperforming conventional methods. This work delivers a valuable rich multi-platform data resource, offering a standardized pipeline for performance benchmarking and a machine learning-based strategy for optimized somatic mutation detection.

bioinformatics↗

bioRxiv2026

Drought Dominates Engineered Lipid Sink Effects on Sorghum Physiology and Carbon Allocation

Determining the environmental limits that govern how engineered metabolic traits control whole-plant carbon (C) allocation is essential for developing climate-resilient bioenergy systems. Oil-enhanced sorghum offers a promising strategy to boost aboveground energy density, yet its system-wide effects on root C investment and soil C delivery, particularly under drought, are largely uncharacterized. Here, we provide the first comprehensive assessment of these interactions using whole-plant 13CO2 continuous labeling, depth-specific sampling of roots and rhizosphere soil, and coordinated measurements of photosynthesis, stomatal anatomy, tissue nitrogen, and root non-structural carbohydrates (NSCs). Under well-watered conditions, oil enhancement produced distinct aboveground physiological changes, including slightly longer stomata, lower stomatal density, significantly higher leaf nitrogen concentration and biomass, and greater leaf 13C enrichment. Importantly, these aboveground changes did not translate into detectable shifts in the allocation of recent photosynthate to belowground pools. In contrast, drought acted as a dominant regulatory factor, reorganizing C flow across both genotypes by suppressing photosynthesis and leaf water status, increasing root nitrogen and NSC reserves, and significantly promoting the retention of recent assimilates in shallow root systems. Depth-specific 13C patterns showed that drought reduced new C incorporation into deep roots while increasing 13C enrichment in deep rhizosphere soil, suggesting that drought altered the balance between C investment in root growth and rhizodeposit C inputs at depth. Ultimately, soil moisture availability was an overriding determinant of belowground C partitioning and vertical C delivery, superseding the influence of the engineered lipid sink. These findings provide new mechanistic insight into the environmental constraints on C flow in an engineered bioenergy crop and identify moisture-driven limitations as the primary bottleneck for translating synthetic metabolic innovations into robust, ecosystem-level C outcomes.

ecology↗

bioRxiv2026

Alignment-free phylogenetic inference via hyperbolic protein language models

Conventional phylogenetic methods rely on multiple sequence alignments which are computationally intensive and often fail for highly divergent lineages. Here, we introduce LucaPhylo, an alignment-free framework that infers evolutionary relationships directly from unaligned sequences. Through a cascaded learning strategy LucaPhylo integrates protein language models with hyperbolic geometry, a representation space naturally suited to hierarchical branching, to capture deep evolutionary constraints without explicit homology matching. Using highly divergent RNA virosphere as a test case, LucaPhylo places unaligned sequences into phylogenetic trees with an accuracy comparable to leading alignment-based tree construction tools, while retaining divergent sequences that conventional pipelines frequently discard. It further enables the integration of divergent viral lineages into phylogenetic trees, thereby expanding the evolutionary landscape of RNA viruses. Together, LucaPhylo establishes an AI-driven, alignment-free paradigm for phylogenetic inference and provides a robust computational foundation for resolving deep evolutionary relationships among RNA viruses and other biological systems.

evolutionary biology↗

bioRxiv2026

Human hindbrain organoids with segment-specific 5-HT neurons for motor modulation

The serotonergic system regulates human mood, reward, and motor control, and its impairment contributes to debilitating psychiatric disorders. Serotonergic neurons (SNs) arise predominantly in the hindbrain and acquire segment-specific identities and functions; however, the lack of human models remains a challenge to understanding the human serotonergic system. Here, we generate human serotonergic-enriched rostral and caudal hindbrain organoids (hsrHOs and hscHOs) that produce segment-specific SNs with regional identities corresponding to rhombomeres r2-r3 and r5-r8, respectively. hsrHOs and hscHOs produce SNs that recapitulate their in vivo counterparts in transcriptomic profiles, function, and ascending or descending projection preferences. Assembly with human neuromusculoskeletal organoids enables modeling of serotonergic modulation of motor control. This system further reveals that deficiency of the schizophrenia risk gene DISC1 disrupts human serotonergic development and impairs motor modulation. Together, our study establishes a human organoid platform for modeling development, circuit function and disease-associated dysfunction of the serotonergic modulatory system.

neuroscience↗

bioRxiv2026

Gene-Specific Analysis of Clonal Hematopoiesis Identifies ASXL1 as a Risk Factor for Lung Cancer

IntroductionClonal hematopoiesis of indeterminate potential (CHIP) is a recognized risk factor for hematologic malignancies, but its contribution to different types of solid cancers remains incompletely defined. MethodsHere, we performed a systematic, gene-specific analysis of CHIP across 19 common solid cancer types using two large population-based cohorts, the UK Biobank and All of Us with Cox proportional hazards models and nested case-control logistic models. ResultsWe demonstrate that the relationship between CHIP and solid tumors is highly cancer-type specific, with lung cancer exhibiting the strongest association. In lung cancer, this association is largely driven by ASXL1-mutant clones. Specifically, high variant allele fraction (high-VAF) ASXL1 conferring a significantly increased risk (hazard ratio = 3.2), and the associations remained robust after adjustment for age, sex, body mass index (BMI), smoking status, and genetic ancestry. Notably, ASXL1 CHIP was substantially enriched among smokers, and its association with lung cancer risk was restricted to ever-smokers, highlighting a key interaction between CHIP and environmental exposure. The enrichment of ASXL1 CHIP in lung cancer was further validated in two independent cancer-only cohorts, including MSK-IMPACT and TCGA. In addition, rare germline variant association analysis revealed that germline variation in ASXL1 had the strongest association with lung cancer susceptibility among all solid tumors. ConclusionsCollectively, our findings support a model in which smoking-associated expansion of ASXL1-mutant clones contributes to lung cancer development and suggest that gene-specific CHIP metrics may enhance risk stratification and early detection strategies.

bioinformatics↗

bioRxiv2026

Genetic Rescue of a Lethal Wasting Mutation (Syb1lew/lew) by Neuron-Specific Expression of ECFP-Syb2

Lethal-wasting (lew) is a spontaneous null mutation in the Syb1/Vamp1 gene encoding the vesicular SNARE protein synaptobrevin 1 (SYB1/VAMP1). Homozygous Syb1lew/lew mice exhibit profound impairment in neuromuscular transmission and die around three weeks after birth. Pathogenic variants in human SYB1 are associated with a variety of disorders such as hereditary spastic ataxia and congenital myasthenic syndrome. Although Syb1 expression is highly enriched in neurons, it has also been reported in non-neuronal tissues, raising the possibility that non-neuronal defects contribute to the lethal phenotype. Here, we tested whether neuronal dysfunction is the primary cause of death in Syb1lew/lew mice and whether the closely related isoform synaptobrevin 2 (Syb2/VAMP2) can substitute for Syb1functin in vivo. We show that neuron-specific expression of ECFP-Syb2 in Syb1lew/lew mice fully rescues lethality in Syb1lew/lewmice, restoring normal growth and motor function. Electrophysiological analyses demonstrate complete recovery of neuromuscular synaptic transmission, including spontaneous and evoked release as well as short-term plasticity. These findings establish that neuronal expression of ECFP-Syb2 is sufficient to prevent the lethal phenotype associated with Syb1 deficiency and demonstrate that Syb2 can functionally replace Syb1 at motor nerve terminals in vivo. Significance StatementNeuron-specific expression of Syb2 fully rescues survival, motor function, and neuromuscular transmission in Syb1lew/lew mice. These findings establish presynaptic Syb1 dysfunction as the primary cause of the lethal phenotype and demonstrate that Syb2 can fully substitute for Syb1 in vivo.

neuroscience↗

bioRxiv2026

Loss of ACTA1 leads to delayed γ-AChR / ε-AChR switch in skeletal muscle in mice

Skeletal muscle actin forms the core structural component of thin filaments, which interact with thick filaments to generate contractile force. In addition to force production, the character of muscle contraction activity itself is thought to provide mechanical cues that influence synaptic development and maturation. In mouse skeletal muscle there is an early post-natal switch from embryonic forms of actin to the adult isoform, ACTA1, which increases both filament stability and force production. Newborn mice deficient for ACTA1 (Acta1-/-), although initially able to breath, move and suckle, develop profound muscle weakness and die during the early neonatal period, despite a compensatory, increase in expression of embryonic actins. We took advantage of this to better understand the response of the neuromuscular junction (NMJ) to a disruption in contractility and activity-dependent signaling during development. Morphological analyses of the diaphragm in Acta1-/- mice revealed that the patterning and formation of the NMJ proceed normally through postnatal day 5 (P5), the day at which pups begin to die. Short-term synaptic plasticity, assessed as the endplate potential (EPP) response to paired-pulse stimulation, was also unchanged, indicating normal presynaptic release of neurotransmitters. In contrast, electrophysiological recordings demonstrated significantly prolonged rise and decay kinetics of miniature and evoked endplate potentials, indicating altered postsynaptic receptor properties. Consistent with these functional changes, quantitative real-time PCR showed a reduced ratio of {varepsilon}- to {gamma}-acetylcholine receptor (AChR) subunit mRNA, reflecting a delay in the developmental switch from embryonic {gamma}-containing to adult {varepsilon}-containing AChRs. Together, these findings indicate that -skeletal actin is dispensable for early NMJ morphogenesis but is required for timely postsynaptic receptor maturation, demonstrating a critical role for muscle contractile activity in coordinating synaptic development at the NMJ. HighlightsO_LISkeletal muscle -actin (ACTA1) is the principal structural component of thin filaments and a key determinant of contractile activity. C_LIO_LIUsing Acta1-/- mice, we show that NMJ patterning and early morphogenesis occur normally despite severe impairment in muscle contractility. C_LIO_LIElectrophysiological analysis of the NMJ shows that presynaptic function remains intact, as evidenced by normal paired-pulse responses. In contrast, postsynaptic maturation is disrupted, with prolonged endplate potential kinetics indicating altered AChR function. C_LIO_LIThis defect is associated with a delayed {gamma}- to {varepsilon}-AChR subunit switch, a key step in postnatal NMJ maturation. These findings identify ACTA1-dependent contractile activity plays a critical role in timely postsynaptic receptor maturation. C_LI

neuroscience↗

bioRxiv2026

Transgenic human dipeptidyl peptidase-4 Syrian hamsters support MERS coronavirus infection and contact transmission

Middle East respiratory syndrome coronavirus (MERS-CoV) is a global health concern due to a high fatality rate associated with human infections and no approved vaccines or therapeutics. While Syrian hamsters are a value animal model for coronavirus research, including SARS-CoV-2, MERS-CoV does not infect wild-type hamsters. Here, we generated transgenic Syrian hamsters expressing human dipeptidyl peptidase-4 (hDPP4), the cellular receptor for MERS-CoV., MERS-CoV replicated efficiently in the respiratory tract tissues of hDPP4 hamsters, causing lethal disease. Treatment with the 3CLpro inhibitor nirmatrelvir significantly reduced viral titers in the lower respiratory tract of infected hDPP4 hamsters. While airborne transmission was not observed, direct contact transmission was observed in all contact hDPP4 hamsters cohoused with infected cage mates. Immunization with purified MERS receptor-binding domain protein reduced virus replication and disease severity but did not prevent direct contact transmission. Collectively, our findings demonstrate that hDPP4 transgenic Syrian hamsters are useful for studying MERS-CoV pathogenesis, transmission, and countermeasure efficacy.

microbiology↗

bioRxiv2026

First evaluation of a human DPP4 transgenic hamster model for MERS-CoV pathogenesis and transmission

MERS-CoV poses a constant pandemic risk, as its viral lineages continue evolving, and zoonotic spillover events could lead to random viral polymorphisms that might lead to human adapted variants. Currently, no small animal model reliably recapitulates both disease progression and transmission dynamics, which are critical aspects for counter-viral measures like vaccine development. Although the Syrian hamster is an optimal animal model for SARS-CoV-2 infection and transmission, it is naturally resistant to MERS-CoV infection. Dipeptidyl peptidase-4 (DPP4) is the functional receptor for MERS-CoV infection, and is highly expressed in human kidney, intestine, liver, and lung tissues. Here, we evaluated the suitability of a human DPP4 (hDPP4) transgenic Syrian hamster model for MERS-CoV research. We used two different MERS-CoV strains (EMC/2012 and D10540/2023) for intranasal inoculation of hamsters. Both strains replicated efficiently, led to comparable severe clinical outcomes, and had similar viral transmission efficiencies. MERS-CoV RNA and nucleoprotein antigen were mainly detected in the brain and the respiratory tract. In summary, we validated a novel hDPP4-transgenic hamster as a suitable model for MERS-CoV infection enabling vaccine and transmission research.

microbiology↗

bioRxiv2026

Efficient and accurate near telomere-to-telomere haplotype reconstruction of diploid genomes

Telomere-to-telomere (T2T) and haplotype-resolved assembly are crucial for understanding eukaryotic genomes. For diploid species, this resolution is critical to uncover allelic variations, inheritance patterns, and functional genomic traits. Current scaffolding methods typically employ either sequence-based or graph-based strategies. Sequence-based approaches rely on proximity signals to yield high contiguity, but underutilize assembly graph information, resulting in more structural errors and chromosomal misassignments. Graph-based methods leverage graph topology for higher accuracy but frequently struggle to achieve chromosome-scale contiguity. However, neither strategy alone can overcome its inherent limitations to simultaneously achieve high contiguity and accuracy. To address these challenges, we introduce HapFold, the first hybrid scaffolding framework that synergistically leverages the complementary strengths of both graph-based and sequence-based approaches. By integrating the topological accuracy of assembly graphs with the proximity-guided contiguity of sequence models, HapFold achieves highly accurate, chromosome-scale or near-T2T haplotype reconstructions for diploid genomes. Compared to existing methods, HapFold achieves superior assembly quality while accelerating computation by an order of magnitude. Furthermore, in the haplotype reconstruction of diploid genomes using standard Oxford Nanopore Technologies simplex reads, HapFold enables the reconstruction of a greater number of near-T2T assemblies. Our approach provides a robust and scalable solution for the high-fidelity reconstruction of haplotype-resolved diploid genomes.

bioinformatics↗

bioRxiv2026

Genos-m: a foundation model for human-associated microbial genomes

Human-associated microbial genomes encode extensive strain-level diversity and niche-specific gene repertoires that are critical to host health. However, these complex sequence features remain difficult to capture using general-purpose DNA foundation models, highlighting the need for dedicated representation learning tailored to the human microbiome. Here, we introduce Genos-m, an open-source foundation model for human-associated microbial genome representation. Genos-m was pretrained on approximately 1.2 trillion nucleotide tokens from a curated microbial genome corpus, including human-associated prokaryotic isolates, high-quality metagenome-assembled genomes (MAGs) and bacteriophages, supplemented with GTDB species-level representative genomes to broaden prokaryotic taxonomic breadth. The model uses a sparsely activated Mixture-of-Experts (MoE) Transformer architecture, with 4.7 billion total parameters, approximately 330 million activated parameters per forward pass and a maximum context length of one million base pairs. We evaluated frozen Genos-m representations across short-sequence and gene-level tasks, biosynthetic gene cluster (BGC)-based regional sequence tasks, whole-genome strain phenotype prediction, and zero-shot transfer on prokaryote-related RNAfitness assays. Across these benchmarks, Genos-m consistently ranked among the leading comparison models, with the best performance in five of eight gene-fitness regression tasks and in BGC type classification. Using sparse autoencoders, we identified sparse features in Genos-m hidden activations that aligned with annotated ORFs, intergenic regions, and tRNA and rRNA loci. In downstream applications, Genos-m-derived genome-informed species representations in-corporated into a human microbiome self-supervised learning model improved colorectal cancer (CRC)-control classification over conventional species-abundance random forest models. Genos-m also generated stable sample-level embeddings from as few as 10,000 metagenomic reads, retaining gut microbial community structure that distinguished geographic origin and aligned with enterotypes defined from full-depth taxonomic profiles. Together, these results support Genos-m as a reusable representation model for microbial genomes and metagenomes, with conclusions bounded by the reported datasets, task definitions and evaluation protocols. Genos-m model weights, inference code, and usage documentation are publicly available on GitHub (https://github.com/BGI-HangzhouAI/Genos-m) and Hugging-Face (https://huggingface.co/BGI-HangzhouAI/Genos-m).

bioinformatics↗

bioRxiv2026

Ferrous Iron Accumulation Is a Hallmark and Therapeutic Vulnerability of Therapy-Induced Senescence

Chemotherapy and radiation reduce tumor burden but leave behind residual cells that survive via therapy-induced senescence (TIS). These cells constitute a latent reservoir fueling recurrence, yet strategies for their selective elimination are lacking. Here, we identify lysosomal ferrous iron accumulation as a conserved hallmark and actionable vulnerability of TIS tumor cells. Across diverse models, senescent tumor cells exhibit marked hypersensitivity to ferroptosis induction. In breast cancer PDX models, sequential ferroptosis induction following chemotherapy significantly delays recurrence, while dual inhibition of GPX4 and FSP1 produces durable, often complete, eradication of residual tumors without overt toxicity. Mechanistically, activation of the TFEB-HO-1 axis in TIS tumor cells drives ferrous iron accumulation, thereby priming cells for ferroptosis. Together, these findings establish ferrous iron accumulation as a defining feature of TIS and position ferroptosis induction as a potent senolytic strategy to eliminate therapy-refractory residual disease. Statement of significanceSenescent tumor cells remaining after treatment can drive cancer recurrence yet remain poorly understood and therapeutically intractable. Here, we identify lysosomal ferrous iron accumulation as a universal hallmark of therapy-induced senescence and demonstrate that ferroptosis induction functions as an effective senolytic strategy. Our findings provide mechanistic and translational support for the "one-two punch" therapeutic paradigm.

cancer biology↗