Biology subjects

Lu, H.

Publications and source records attributed to Lu, H..

At least 73 records · Page 4Linked to original sources

bioRxiv2025

High-density theta burst stimulation (hdTBS) at 100 Hz triples the aftereffects of the conventional intermittent TBS

Slice electrophysiological studies have experimentally demonstrated that theta burst stimulation, consisting of electrical pulses delivered at 10 ms (100 Hz) inter-pulse intervals, optimally induces long-term potentiation in the hippocampus. Inspired by this observation, a novel transcranial magnetic stimulation (TMS) paradigm, 100 Hz high-density theta burst stimulation (100 Hz hdTBS), is presented. This paradigm delivers 6 pulses per burst with an inter-pulse interval of 10 ms - doubling the pulse frequency and total pulse count of the conventional intermittent TBS (iTBS). The effect of this new paradigm was studied in the motor cortex of awake rats using a rat-specific focal TMS coil and a hdTBS stimulator developed in house. Results reveal that 100 Hz hdTBS triples the after-effects of conventional iTBS. In a separate group of animals that received two consecutive iTBS session back-to-back (prolonged iTBS), we observed an inhibitory effect. Since that prolonged iTBS matches the total pulse count of 100 Hz hdTBS but produced opposite after-effects, our results underscore the critical roles of the temporal structure of TMS pulses--not merely the total number of pulses--in driving neuroplasticity. This new paradigm has the potential to significantly enhance therapeutic efficacy if confirmed to be safe and effective in humans.

neuroscience↗

bioRxiv2025

Universal principles of protein resource allocation among cellular organelles revealed by yeast large-scale absolute quantitative proteomics

The presence of organelles is a hallmark distinguishing eukarya from bacteria and archaea, and this culminates in compartmentalization of cellular metabolism and subsequent metabolic specialization. Here we established a dataset encompassing over 300 absolute quantitative proteomes, the largest to date, across two yeast species under diverse experimental conditions. Leveraging big data analysis, formula fitting, and machine learning models, quantitative correlations among protein abundance, organelle-level resource distribution, and cellular phenotypes were elucidated at a system level. We found that protein resources always exhibit robust and precise distribution at the organelle level across distinct conditions. Specifically, at high specific growth rates, the protein mass fraction from some main organelles, i.e., peroxisome and nucleus, is consistently reduced to offset the increasing protein resource demand from the ribosome. Meanwhile, we found that the nutrition limitation could induce resource recycling by upregulating protein resources within the vacuole and lipid droplets to sustain stress adaptation. Importantly, our integrative analysis demonstrates that protein mass fraction from less than 4 organelles (e.g., nucleus and ribosomes) can accurately predict diverse yeast physiological parameters (e.g., specific growth rate, oxygen uptake rate), and a core set of 37 proteins could predict resource allocation among 24 main organelles and sub-organelles with high accuracy (average R^2 > 0.9). Finally, we found organelle resource allocation reflects the divergence of yeast species. For example, anaerobic conditions and respiratory suppression have less influence on Crabtree-positive yeast, i.e., Saccharomyces cerevisiae, with respect to organelle resource allocation but have a larger effect on the Crabtree-negative yeast Issatchenkia orientalis, thus suggesting that cellular resources have facilitated adaptive evolution. In summary, the high-quality, genome-scale quantitative proteomic dataset for yeast species offers an unprecedented opportunity for understanding the basic principles underlying resource allocation at the organelle level, laying theoretical foundations for precision engineering of cell factories in synthetic biology. The resource used in this study is available at https://yeast-proteome-database.streamlit.app/.

systems biology↗

bioRxiv2025

Conformational Dynamics of the Free MAX Protein Revealed by Well-Tempered Metadynamics

One of most relevant challenges in tumorigenesis is the association of MYC and MAX proteins, whose related cancers remain undrugged. In particular, the disordered regions shown by those oncogenes make their structural characterization and the development of new drugs a truly hard task. To address these challenges, we employed the enhanced-sampling well-tempered metadynamics method to systematically explore the conformational space of the unbound MAX protein. Our results revealed, for the first time, a well-defined and thermodynamically favorable conformation of monomeric MAX. This is a remarkable finding, as it demonstrates that regions of MAX previously considered persistently disordered are capable of adopting stable, folded structures under specific conditions. Moreover, our findings also suggest that the meta-stable structural motifs observed in this work may harbor druggable sites, particularly relevant for strategies aiming to target MAX directly or to disrupt its interaction with MYC, thereby modulating oncogenic signaling pathways. The present study establishes a new structural framework for understanding the dynamics of MAX and provides a foundation for future structure-based drug design targeting the MYC/MAX axis. Finally, our work offers a strategic blueprint for investigating similarly challenging drug targets. TOC Graphic O_FIG O_LINKSMALLFIG WIDTH=200 HEIGHT=111 SRC="FIGDIR/small/673476v3_ufig1.gif" ALT="Figure 1"> View larger version (29K): org.highwire.dtl.DTLVardef@18ea239org.highwire.dtl.DTLVardef@13c8b8eorg.highwire.dtl.DTLVardef@1ee414corg.highwire.dtl.DTLVardef@10114c1_HPS_FORMAT_FIGEXP M_FIG C_FIG

biophysics↗

bioRxiv2025

Adaptive mutations at lysine residues of PRRSV nsp12 enable evasion of host proteasomal and selective autophagic degradation

Ubiquitin signaling in viral infections can either enhance immune responses and degrade viral proteins or be exploited by viruses to target host antiviral factors. Porcine reproductive and respiratory syndrome virus (PRRSV) poses a significant threat to the pig industry, with its non-structural proteins (nsps) critical for virulence and replication. Here, we investigated the role of ubiquitination in the stability of PRRSV-encoded nsps and found that nsp12 was specifically degraded via the ubiquitin-proteasome system. Mechanistically, nsp12 underwent K48- and K63-linked polyubiquitination at lysine residues 89, 91, 127, and 130. RNF114 served as an E3 ubiquitin ligase for nsp12, with its enzymatic activity essential for both nsp12 degradation and viral replication. Additionally, nsp12 underwent ubiquitin-dependent selective autophagy through receptor-mediated recognition, wherein NBR1, SQSTM1, and NDP52 bridged its interaction with LC3 for autophagic degradation. Evolutionary analyses revealed that PRRSV nsp12 acquired non-lysine residues at positions 89, 127, and 130 during viral adaptation. Correspondingly, recombinant PRRSV strains carrying the K91/127/130R mutations within nsp12 exhibited enhanced replication, while a revertant strain with the R89K mutation in nsp12 showed attenuated infectivity. Mass spectrometry analysis further identified significant enrichment of ubiquitination-related modifications among nsp12-interacting proteins. These findings provide valuable insights for anti-PRRSV drug design and highlight the challenge posed by adaptive mutations in viral proteins to the swine industry. Author SummaryUbiquitination plays crucial roles in both proteasomal degradation and selective autophagy during viral infections, yet its impact on the stability of porcine reproductive and respiratory syndrome virus (PRRSV) nonstructural proteins (nsps) remains unexplored. Here, we found that nsp12 was targeted for proteasomal degradation through K48/K63-linked polyubiquitination at lysine residues 89, 91, 127, and 130, which requires the involvement of E3 ubiquitin ligase RNF114. Furthermore, nsp12 underwent receptor-mediated selective autophagic degradation through the action of NBR1, SQSTM1, and NDP52. Using reverse genetics technology, mutations of lysine to arginine in PRRSV nsp12 enhanced viral replication, whereas the reverse mutations reduced its infectivity. Our findings demonstrate that PRRSV evades host degradation by acquiring adaptive mutations within nsp12, which counteract ubiquitin-dependent clearance thus enhancing viral fitness. This novel mechanism illustrates a key viral immune evasion strategy and underlines the challenge in controlling PRRSV.

microbiology↗

bioRxiv2025

EnzymeTuning: a Deep-learning-based Toolbox for Optimizing Enzyme-constrained Metabolic Modeling with Enhanced Proteome Abundance Prediction

AbstractThe accuracy of enzyme kinetic parameters, particularly the turnover number (kcat), is critical for the predictive power of enzyme-constrained genome-scale metabolic models (ecGEMs). However, current kinetic datasets remain sparse and often fail to capture in vivo enzyme behavior, compromising model predictive capacity. To address these challenges, we developed EnzymeTuning----a generative adversarial network (GAN)-based framework for the global kcat optimization. This approach significantly enhances both the accuracy and proteome-level coverage of ecGEM predictions. Moreover, by incorporating literature-derived protein degradation constants (kdeg), we inferred the protein synthesis rate and systematically evaluated their impact on model performance. The framework was validated across distinct yeast species, including Saccharomyces cerevisiae, Kluyveromyces lactis, Kluyveromyces marxianus, and Yarrowia lipolytica, demonstrating its generalizability. Further, we found that EnzymeTuning facilitates the identification of context-specific enzyme usage patterns and adaptive catalytic resource allocation under varying carbon-to-nitrogen (C/N) ratios, showcasing the substantial potential of our toolbox for integrative omics analysis. Overall, EnzymeTuning provides a robust and scalable solution for refining kinetic parameters in ecGEMs, thereby promoting the wide applications of these computational models in systems and synthetic biological studies.

systems biology↗

bioRxiv2025

scYeast: a Biological-knowledge-guided Foundation Model on Yeast Single-Cell Transcriptomics

Though large-scale pre-trained models are vital for foundational cell modeling, most focus on human or mouse systems, neglecting model organisms like yeast (Saccharomyces cerevisiae) and often failing to use existing biological prior knowledge effectively. To address this, we present scYeast, the first foundational cell model tailored for yeast that seamlessly embeds biological priors. scYeast employs a novel asymmetric parallel architecture to infuse transcriptional regulatory information directly into the Transformers attention mechanism, leveraging established biological knowledge during training. Pre-trained on large-scale yeast single-cell transcriptomics data, scYeast demonstrates strong generalization and biological interpretability. It excels in zero-shot tasks, such as inferring regulatory relationships and identifying critical cell states. After fine-tuning, scYeast performs exceptionally across a diverse set of tasks from cell type classification to predicting growth doubling time and gene perturbation response. Additionally, using transfer learning, scYeast can be adapted to other omics datasets, such as proteomics, thus broadening its utility in big data analysis. Overall, scYeast is a powerful tool for yeast single-cell biology research and sets a new standard for integrating foundational models with biological prior knowledge, dramatically accelerating the pace of discovery in yeast synthetic and systems biology and providing a replicable framework for other organisms.

bioinformatics↗

bioRxiv2025

Cis-delivering releasable IL-15 superagonist enhances antitumor immunity in cold tumors by invigorating preexisting CD25+CD8+ T cells

Antibody-mediated cis-delivery and trans-delivery both direct cytokines to tumors and have been extensively investigated in clinical trials. However, a comparative analysis of their differential effects on cytokine activity is still lacking. In this study, we initially verify that cis-delivery of cytokine demonstrates a markedly stronger antitumor effect than trans-delivery, but it also exhibits certain drawbacks, including severe toxicity, insufficient activation of CD25+CD8+ T cells, and enhanced stimulation of intratumoral regulatory T cells (Tregs). To further address these issues, we developed a conditionally releasable and cis-delivering IL-15 immunocytokine (termed PMIS), which can selectively release a free IL-15 superagonist within tumors rather than immobilizing IL-15 to PD-1+ Tregs, thereby potently stimulating CD25+CD8+ T cells. Mice treated with PMIS showed significantly reduced systemic toxicities while achieving notably stronger antitumor effects. Administered either alone or in combination with other therapies, PMIS exhibits great potential for inhibiting orthotopic cold tumor and its metastases. Mechanistically, the significant activation of the pre-existing intratumoral CD25+CD8+ T cells and the improved CD25+CD8/Treg ratio contribute to the enhanced antitumor response of PMIS. These findings underscore the indispensable role of CD25+CD8+ T cells in cis-delivering IL-15 and provide a promising strategy for overcoming resistance to therapies and effectively controlling advanced cold tumors.

immunology↗

bioRxiv2025

A Reproducible Supermicrosurgical Model of Cervical Lymphatic-Venous Anastomosis in a Transgenic Alzheimer's Mouse: Technical Translation, Proficiency, and a Validated Learning Curve

BackgroundCervical Lymphatic-Venous Anastomosis(LVA) holds therapeutic potential for Alzheimer disease (AD). However, a validated AD animal model for preclinical investigation remains lacking. This study aimed to establish and standardize LNVA, as a form of lymphatic-venous Anastomosis, in a murine AD model; to quantitatively assess its surgical learning curve; and to create a platform for mechanistic exploration and translational research. MethodsWe established and standardized a deep cervical lymph node-to-vein anastomosis (DC-LNVA), a specific type of lymphatic-venous anastomosis (LVA), in rodents. A two-stage training program was applied to a junior surgeon: 60 DC-LNVA procedures were performed in 30 small Sprague-Dawley rats (50-60g), followed by 20 bilateral anastomoses in 10 male 5XFAD mice. Surgical proficiency was evaluated using procedure time, UWOMSA scoring, and CUSUM analysis. Anatomical parameters were assessed with ImageJ. Anastomotic patency was confirmed via ICG lymphangiography. Intranasal Evans Blue was used to trace cranial distribution pathways. ResultsSD rats had significantly larger external jugular veins (0.376 {+/-} 0.013 mm vs. 0.228 {+/-} 0.011 mm) and DCLNs (1.359 {+/-} 0.084 mm2 vs. 0.333 {+/-} 0.022 mm2) than mice (P < 0.0001). Proficiency was achieved after 27 anastomoses in rats and 10 in mice. ICG confirmed 100% patency in 9 surviving 5XFAD mice. Evans Blue suggested a nasal-cranial pathway for tracer distribution. ConclusionWe developed and validated a cervical lymphatic-venous bypass (LVA) in transgenic AD mice. This platform complements existing loss-of-function approaches and offers a foundation for mechanistic and translational research into brain lymphatic clearance.

animal behavior and cognition↗

bioRxiv2025

Endothelial cell-specific DNA methylation alterations identified in breast cancer with epigenetic deconvolution

DNA methylation alterations are epigenetic modifications that regulate gene expression and contribute to carcinogenesis. Technical challenges and high costs limit the use of single-cell sequencing approaches to measure and understand cell-specific DNA methylation alterations in the tumor microenvironment (TME). Using cell type deconvolution (HiTIMED) and an interaction testing framework (CellDMC), we identified DNA methylation alterations in tumor endothelial cells (TECs) from measures in bulk biospecimens. Alterations in TECs contribute to abnormal angiogenesis, resulting in disorganized, leaky blood vessels that can hinder drug delivery and immune access. We hypothesize that TECs in the TME exhibit distinct DNA methylation patterns compared to normal endothelial cells, driving phenotypic changes that support tumor growth and vascular dysfunction. We used genome-scale DNA methylation data and a testing and validation set design (tumor, n=1,080; nontumor, n=415) to identify TEC-specific DNA methylation alterations in breast cancer. Between cohorts, we validated 4,562 TEC-specific CpGs (4,163 hypermethylated and 399 hypomethylated), and found that many map to genes regulating angiogenesis and endothelial function. Integration with mRNA data from tumor samples in both discovery and validation sets revealed genes whose expression correlates with TEC-specific DNA methylation patterns and endothelial cell proportions, highlighting potentially novel therapeutic targets. This study demonstrates the utility of cellular deconvolution to uncover cell lineage-specific epigenetic changes in cancer using bulk specimen, and identifies TEC-specific methylation alterations as novel potential targets for therapeutic intervention.

cancer biology↗

bioRxiv2025

ASCENT: Annotation-free Self-supervised Contrastive Embeddings for 3D Neuron Tracking in Fluorescence Microscopy

Tracking individual neurons in dense 3D microscopy recordings of fast-moving, deforming organisms is a critical challenge in neuroscience. This task is often hindered by complex tissue motion and the need for laborious manual annotation. Here, we introduce ASCENT, an annotation-free computational framework that learns robust representations of neuronal identity that are invariant to tissue deformation. ASCENT trains a Neuron Embedding Transformer (NETr) through a self-supervised contrastive scheme on augmented data that mimics tissue deformations. NETr generates highly discriminative embeddings for each neuron by combining visual appearance with positional information, refined by the context of all other neurons in the frame. On a challenging benchmark C. elegans datasets, ASCENT achieves state-of-the-art tracking accuracy, surpassing supervised methods. The method is robust to image noise, highly data-efficient, and generalizes across different imaging conditions. By eliminating the annotation bottleneck, ASCENT provides a practical and scalable solution for the robust analysis of whole-brain neural dynamics in behaving animals.

bioengineering↗

bioRxiv2025

High speed functional imaging with a microfluidics-compatible open-top light-sheet microscope enabled by model predictive control of a tunable lens

1Functional fluorescence imaging of small, transparent organisms has proven to be a powerful tool for understanding the development and function of intact nervous systems at the cellular level, particularly when combined with microfluidic tools that enable precise manipulation of sensory cues. Light-sheet fluorescence microscopy has a number of advantages for functional imaging, including efficient axial sectioning, reduced photobleaching, and high speed compared to confocal methods, but many configurations with ideal properties for efficient functional imaging place constraints on sample access and geometry that preclude their use with microfluidics. We present an open-top light-sheet microscope that uses a simplified inverted water immersion interface coupled with tunable lens remote focusing to achieve high-speed, multichannel 3D imaging of specimens in conventional microfluidics. We use model predictive control to efficiently optimize drive signals for an electrically tunable lens, providing reliable, camera-limited scanning of the imaging volume. We then demonstrate the utility of this approach by recording calcium activity from C. elegans at volume rates up to 20 Hz, including both whole-brain response to chemical stimulation and compartmentalized dendritic response of the PVD neuron to mechanical stimulation. Our approaches are flexible and inexpensive to implement, and could be adapted to improve performance and sample compatibility for a wide range of imaging techniques.

neuroscience↗

bioRxiv2025

AlphaGEM Enables Precise Genome-Scale Metabolic Modelling by Integrating Protein Structure Alignment with deep-learning-based Dark Metabolism Mining

Constructing high-quality genome-scale metabolic models (GEMs) for less-studied species remains challenging. To address this, we developed AlphaGEM, a versatile toolbox leveraging proteome-scale structural alignment and deep-learning-based predictions for efficient genomic mining to generate GEMs ready for applications. Our findings show that the structural alignment or protein-language-model-based prediction (i.e., PLMSearch), could identify more homologous protein relationships than sequence-blast-based alignment, contributing to the accurate profiling of metabolism from target organisms. Additionally, AlphaGEM encompasses an ensemble procedure empowered by multiple deep learning toolboxes to effectively mine the dark metabolic functions encoded by nonhomologous proteins, significantly expanding species-specific metabolic networks. We demonstrate AlphaGEMs accuracy by building GEMs for eukaryotes (e.g., S. pombe, C. albicans) and prokaryotes (e.g., K. pneumoniae, B. subtilis), achieving predictions comparable to manually curated models while outperforming existing tools. AlphaGEM also successfully reconstructs GEMs for M. musculus and C. griseus, showcasing its great potential for uncovering dark metabolism in complex mammals. Lastly, we demonstrate that AlphaGEM could facilitate the automatic GEMs reconstruction for 332 distinct yeast species with high prediction fidelity. In conclusion, AlphaGEM provides unprecedented opportunities for the precise, rapid construction of GEMs across diverse domains, which sets a solid foundation for universal functional analysis of non-model organisms having genome sequences available.

bioinformatics↗

bioRxiv2025

Maturation of Dorsal Association Tracts during Preadolescence Links to Concurrent and Future Cognitive Performance and Transdiagnostic Psychopathology

Many psychiatric disorders begin during adolescence, coinciding with the rapid development of brain white matter (WM). However, it remains unclear whether deviations from normal WM maturation during this age period contribute to the development of psychopathology. In this study, we developed and validated normative models of brain age based on specific WM tracts using three large-scale developmental datasets (a total of [~]10,000 subjects). We found that tract-specific deviations in WM development of association and limbic/subcortical systems were linked to concurrent cognition and psychopathology. The spatial pattern of the association system aligned closely with distributions of high-order brain networks, and with mitochondrial content and respiratory capacity. The maturation of the association system contributed significantly to better cognitive performance assessed two or three years later. Importantly, delayed WM development especially in dorsal association tracts predicted psychiatric disorders across diagnoses and disorder onset over a 2-year follow-up. By identifying tract-specific WM development during preadolescence as a predictor of cognitive capacity and psychiatric disorder risks, this study provides a valuable framework for tracking individualized brain maturation and understanding the neurobiological underpinnings of cognitive performance and transdiagnostic psychopathology.

neuroscience↗

bioRxiv2025

A Novel FNDC1-NAMPT-NAD axis is Implicated in Small and Large-vessel Arterial Disease and Drives Vascular Calcification

Vascular calcification represents a convergent pathological feature of diverse cardiovascular diseases, yet the upstream molecular programs orchestrating this process remain poorly defined. Here, we uncover fibronectin type III domain-containing 1 (FNDC1) as a previously unrecognized regulator of vascular calcification across both microvascular and macrovascular beds. Integrative transcriptomic profiling of human calciphylaxis lesions and atherosclerotic coronaries identified FNDC1 as one of the most significantly upregulated genes. In primary human vascular smooth muscle cells, FNDC1 drove osteogenic phenotype switch and vascular calcification through activation of PI3K/AKT signaling and metabolic reprogramming. Mechanistically, FNDC1 directly binds to nicotinamide phosphoribosyltransferase (NAMPT) resulting in elevated intracellular NAD levels, thus coupling vascular signaling to control of NAD biosynthesis. In murine models, genetic deletion of Fndc1 or pharmacologic inhibition of NAMPT suppressed arterial calcification and prolonged survival. Clinically, circulating FNDC1 levels were elevated in patients with both calciphylaxis and coronary artery disease and independently predicted cardiovascular risk in 42,687 UK Biobank participants. Together, these findings establish FNDC1 as a central mediator of vascular pathology and highlight the FNDC1- NAMPT-NAD+ axis as a promising target for therapeutic intervention.

cell biology↗

bioRxiv2025

Divergent Neurobehavioral Effects of CFTR Modulators Elexacaftor and Ivacaftor in Mice

Recent advances in cystic fibrosis transmembrane conductance regulator (CFTR) modulator combination therapies have markedly improved survival and quality of life for people with cystic fibrosis (CF). Among these, the elexacaftor-tezacaftor-ivacaftor combination (Trikafta) and its updated formulation, vanzacaftor-tezacaftor-deutivacaftor (Alyftrek), represent major therapeutic milestones. However, emerging reports suggest that Trikafta may contribute to anxiety and depression in some people with CF. In this study, we investigated the neurobehavioral effects of elexacaftor and ivacaftor in mice. Acute administration of elexacaftor elicited anxiety-like behavior, while ivacaftor induced depressive-like behavior. Additionally, we confirmed the presence of Cftr mRNA in the amygdala and hippocampus, brain regions implicated in anxiety and depression. These findings provide preclinical support for the reported mental health side effects experienced by some people with CF and offer new insights into CFTR function within the central nervous system, potentially guiding the development of future CF therapies with improved neuropsychiatric profiles.

physiology↗

bioRxiv2025

EZ-FRCNN: A Fast, Accessible and Robust Deep Learning Package for Object Detection Applications from Ethology to Cell Biology

Advances in high-throughput imaging and experimental automation have dramatically increased the scale of biological datasets, creating a growing need for tools that can efficiently identify and localize features in complex image data. Although deep learning has transformed image analysis, methods such as region-based convolutional neural networks remain underutilized in biology due to technical barriers such as coding requirements and reliance on cloud infrastructure. We present EZ-FRCNN, a locally hosted, user-friendly package that enables the accessible and scalable application of object detection to biological datasets. Through graphical and scriptable interfaces, users can annotate data, train models, and perform inference entirely offline. We demonstrate its utility in detecting cell phenotypes for large-scale screening, enabling the first label-free tracking of grinder motion in freely moving C. elegans to quantify feeding dynamics, and identifying animals in naturalistic environments for ecological field studies. These once-infeasible analyses now enable rapid screening of cell therapies, investigation of internal state-behavior coupling without immobilization or genetic modification, and efficient wildlife tracking with minimal computational cost. Together, these examples demonstrate how accessible tools like EZ-FRCNN can drive new biological discoveries in both laboratory and field environments.

bioengineering↗

bioRxiv2025

Euktect: Enhanced Eukaryotic Sequence Detection and Classification in Metagenomes via the DNA Language Model

The current taxonomy classification of DNA sequences in metagenomics primarily relies on alignment against reference databases. However, the eukaryotic species in genome databases are not sufficient as there exist numerous unculturable eukaryotes inside metagenomes. These limitations hinder functional and evolutionary analyses of eukaryotes across distinct environmental samples. To overcome these limitations, we created Euktect, a deep-learning-based toolbox for reliable, alignment-free classification of eukaryotic DNA sequences across different phylogenetic levels from metagenome datasets. Euktect achieves high accuracy in extraction of eukaryotic sequences longer than 500bp from the assembled contigs of metagenomes, significantly outperforming existing methods. Furthermore, we developed an algorithm that integrates Euktects predictions with existing tools to refine metagenome-assembled genomes (MAGs), substantially increasing the yield of high-quality eukaryotic MAGs for the downstream analysis. Beyond eukaryotic detection, Euktect incorporates two specialized models: a high-precision classifier for fungal phyla, and a hierarchical classifier that accurately identifies sequences from specific fungal genera (e.g., Candida) in diverse metagenomic samples. Significantly, this framework enables prediction of host disease status (e.g., inflammatory bowel disease) by linking eukaryotic sequence identification to clinical phenotypes through machine learning models. Collectively, Euktect enables accurate reconstruction and functional annotation of eukaryotic genomes from metagenomes at large scale, thereby empowering the comprehensive utilization of sequenced eukaryotic species for downstream evolutionary and clinical research.

bioinformatics↗

bioRxiv2025

Development of a replication-defective MPXV platform for fundamental and therapeutic research

The recent global outbreaks of mpox highlight the urgent need for both fundamental research and antiviral development. However, studying monkeypox virus (MPXV), with its large and complex genome, remains challenging due to the requirement for high-containment facilities. Here, we describe a novel strategy for de novo assembly of MPXV clade IIb genomes in bacterial artificial chromosomes using transformation-associated recombination cloning. Leveraging CRISPR-Cas9 and Lambda Red recombination, we engineered replication-defective MPXV particles with dual deletions of OPG96 (M2R) and OPG158 (A32.5L)--genes essential for virion assembly, that are capable of recapitulating key stages of the viral life cycle. Our work demonstrates the utility of replication-defective MPXV particles as a reliable platform for high-throughput antiviral discovery, offering significant advantages for both fundamental virology studies and therapeutic development against orthopoxviruses.

microbiology↗