bioRxiv Science⌕ Search

Biology subjects

Jiang, X.

Publications and source records attributed to Jiang, X..

At least 127 records · Page 7Linked to original sources

Prg4+ fibro-adipogenic progenitors in muscle are crucial for bone fracture repair

Clinically, compromised fracture healing often occurs at sites with less muscle coverage and muscle flaps can provide the necessary healing environment for appropriate healing in severe bone loss. However, the underlying mechanisms are largely unknown. Here, we established a mouse reporter model for studying muscle cell contribution to bone fracture repair. Analyzing skeletal muscle scRNA-seq datasets revealed that Prg4 marks a fibro-adipogenic progenitor (FAP) subpopulation. In mice, Prg4+ cells were specifically located in the skeletal muscle, but not at the periosteum or inside cortical bone. These cells expressed FAP markers, responded to muscle injury, and became periosteal cells under normal and muscle injury conditions. Fracture fragmented muscle fibers, rapidly expanded Prg4+ FAPs at the injury site and promoted their migration into the fracture site. Later, they gave rise to many chondrocytes, osteoblasts, and osteocytes in the outer periphery of callus next to muscle. In repaired bones, the descendants of Prg4+ FAPs were detected as mesenchymal progenitors in the periosteum and osteocytes at the prior fracture site. A second fracture activated those cells and stimulated them to become osteoblasts in the inner part of callus. Importantly, ablation of Prg4+ FAPs impaired fracture healing and functional repair. In an intramembranous bone injury model (drill-hole), Prg4+ FAPs became periosteal cells, but their contribution to bone defect repair was significantly less than in fractures. Taken together, we demonstrate the critical role of FAPs in endochondral bone repair and uncover a novel mechanism by which mesenchymal progenitors transform from muscle to cortical bone. Significance statementFracture healing is often impaired in areas with less muscle coverage, but the mechanisms behind this are poorly understood. In this work, we uncovered a critical role of specific muscle cells, known as Prg4+ fibro-adipogenic progenitors (FAPs), in bone regeneration. Using mouse models, we show that these muscle-resident mesenchymal progenitors expand after bone injury, migrate to the fracture site, transform into various bone cells in the fracture callus, and eventually become bone surface progenitors after the bone is healed. This finding highlights the importance of muscle cells in endochondral fracture repair and reveals a novel crosstalk mechanism between muscle and bone tissues. Future studies targeting these muscle progenitor cells could develop new treatments for delayed and nonunion fractures.

cell biology↗

Divergent Minimum Inhibitory Concentrations of Clarithromycin and Azithromycin Against Clinical Isolates of Mycobacterium abscessus Complex

Current treatment guidelines for Mycobacterium abscessus complex (MAB) infections recommend macrolides (clarithromycin/CAM and azithromycin/AZM) as cornerstone therapies, considering them clinically interchangeable without distinction. This study evaluated the comparative in vitro activity of clarithromycin (CAM) and azithromycin (AZM) against 146 clinical isolates of Mycobacterium abscessus complex organisms. MIC distributions revealed significantly higher resistance to AZM (94.5% resistant at [&ge;]8 g/mL) compared to CAM (77.4% resistant at [&ge;]8 g/mL), with only 4.8% of isolates susceptible to both drugs. Notably, 26 strains (17.8% of total) showed AZM resistance despite being non-resistant to CAM (MIC <8 g/mL) among 146 clinical isolates. Among the 33 CAM-non-resistant isolates (including susceptible and intermediate strains), 78.8% (26/33) were AZM-resistant, demonstrating frequent discordance in susceptibility profiles. These findings contrast with current treatment guidelines that recommend using CAM and AZM interchangeably without distinction. Subspecies analysis showed three distinct susceptibility profiles: isolates susceptible to both antibiotics (n=3, M. massiliense), isolates resistant to both antibiotics (n=3, M. massiliense), and M. bolletii with CAM-susceptible/AZM-resistant phenotypes (n=2). This differential resistance pattern suggests that while both drugs face substantial resistance challenges, CAM retains meaningful activity against certain strains that are AZM-resistant--particularly in M. bolletii. These findings indicate that AZM and CAM should not be used interchangeably to treat MAB infections and emphasize the need to prioritize CAM over AZM in treatment regimens.

microbiology↗

Bridging 3D Molecular Structures and Artificial Intelligence by a Conformation Description Language

Artificial intelligence, particularly language models (LMs), is reshaping research paradigms across scientific domains. In the fields of chemistry and pharmacy, chemical language models (CLMs) have achieved remarkable success in two-dimensional (2D) molecular modeling tasks by leveraging one-dimensional (1D) representations of molecules, such as SMILES and SELFIES. However, extending these successes to three-dimensional (3D) molecular modeling remains a significant challenge, largely due to the absence of effective 1D representations for capturing 3D molecular structures. To address this gap, we introduce ConfSeq, a novel molecular conformation description language that integrates SMILES with internal coordinates including dihedral angles, bond angles, and pseudo-chirality. This design naturally ensures SE(3) invariance, while preserving the human readability and conciseness characteristic of SMILES. ConfSeq enables the reformulation of a range of 3D molecular modeling tasks, such as molecular conformation prediction, 3D molecular generation, and 3D molecular representation, into sequence modeling problems. Then, by simply employing a standard Transformer architecture, we achieve state-of-the-art performance on various benchmark sets. Furthermore, compared to widely used diffusion-based approaches in 3D molecular modeling, the ConfSeq-based method offers unique advantages in inference efficiency, generation controllability, and enables scoring of generated molecules. We believe that ConfSeq can serve as a foundational tool, advancing the development of sequence-based 3D molecular modeling methods.

bioinformatics↗

An embryonic artery-forming niche reactivates in pulmonary arterial hypertension

Developmental programs that orchestrate cell fate and tissue architecture during organogenesis can cause disease when reactivated in adults. Here we identify a population of endothelial cells (ECs) defined by the pioneer factor early B cell factor 1 (EBF1) that controls both pulmonary artery (PA) morphogenesis and pulmonary arterial hypertension (PAH). During embryonic development, the PA emerges from an endothelial niche harbored within the vascular plexus, where Aplnr+ endothelial progenitors give rise to both arterial ECs and EBF1+ ECs. Rather than directly incorporating into the PA intima, EBF1+ ECs control the branching and maturation of the PA tree through paracrine vasculotrophic signals that direct plexus expansion, arterialization, and mural cell recruitment. Although essential for development, most EBF1+ ECs disappear upon completion of PA morphogenesis. In adult PAH, vascular injury reactivates this developmental program: normally quiescent general capillary ECs re-enter the cell cycle and differentiate into arterial ECs and EBF1+ ECs, reconstituting this artery-forming niche in a maladaptive reprise. Unlike their transient embryonic counterparts, EBF1+ ECs persist within neointimal lesions and express vasculotrophic signals associated with pathological PA remodeling. Capillary-restricted Ebf1 induction combined with endothelial injury recapitulates this program, driving neo-arterialization, neointimal formation, and severe PAH. Conversely, endothelial-targeted AAV-mediated Ebf1 knockdown achieves complete protection in a preclinical model of disease, demonstrating that endothelial EBF1 is necessary for PAH pathogenesis. These findings demonstrate that reactivation and persistence of a transient embryonic artery-forming niche in adulthood can promote pathological vascular remodeling and PAH.

developmental biology↗

Robust Bayesian Integrative Modeling of Single Cell and Spatially Resolved Transcriptomics Data

A recent technology breakthrough in spatially resolved transcriptomics (SRT) has enabled comprehensive molecular characterization at the cellular level while preserving spatial information. However, many SRT technologies (e.g., spatial transcriptomics) cannot achieve single-cell resolution. Instead, they measure the average gene expression of a mixture of cells within a spot arranged on a lattice. Here, we introduce iMOSCATO, a fully Bayesian model that integrates single-cell RNA sequencing (scRNA-seq) data and SRT data to simultaneously decompose cell-type compositions of spots and identify the underlying spatial domains. We incorporate the lattice structure by employing a Markov random field prior, improving the accuracy of both cell-type deconvolution and spatial domain detection. Moreover, we use a zero-inflated Dirichlet distribution to capture cell-type sparsity. Finally, iMOSCATO shows competitive performance in accuracy compared to existing methods in both simulation studies and two real data applications.

bioinformatics↗

Large-scale single-neuron recording using the uFINE array in human cortex

Monitoring neural population activity at single-cell resolution is essential for driving fundamental research and clinical innovations. However, translating advanced recording techniques from animal models to humans remains a significant challenge. Flexible neural electrodes have recently emerged as powerful tools for large-scale single-unit recordings due to their superior biocompatibility and high recording density. Here, we demonstrate reliable, high-density single-unit recordings during intraoperative procedures in human patients using ultra-Flexible Implantable Neural Electrode (uFINE) arrays. The uFINE array exhibited sufficient mechanical robustness to maintain structural integrity throughout surgical operations. We successfully recorded 616 single units from 10 patients, with up to 135 single units simultaneously recorded. The flexibility of uFINE array minimized signal disturbances from brain pulsations, enabling stable and continuous single-unit detection. Stimulus and response tuning were observed at the level of individual neurons in awake patients. This uFINE-based recording approach offers unique opportunities to investigate human-specific cognitive functions and develop next-generation brain-machine interfaces.

bioengineering↗

Small non-coding RNAs encapsulating mammalian cells fuel innate immunity

Cell surface RNAs, notably glycoRNAs, have been reported, yet the exact surface RNA compositions in different cell types remain unclear. Here, we introduce a comprehensive suite of methodologies for imaging, profiling, quantifying, and exploring the biological functions of specific surface RNAs. Utilizing these techniques, we have identified diverse non-coding RNAs present on mammalian cell surfaces. We confirm the membrane anchorage and quantify the abundance of several representative RNAs on human primary cells. Notably, we discover a significant prevalence of Y RNAs on the surfaces of human monocytes and B cells. We find that these Y RNAs on human monocyte surfaces enrich extracellular histones, regulating interleukin-6 (IL-6) gene expression and subsequent protein secretion upon histone stimulation via NF-{kappa}B and AP-1 activation. Our study not only presents effective approaches for investigating surface RNAs, but also uncovers a previously unrecognized immune activation pathway mediated by surface Y RNAs on monocytes. In briefA comprehensive profiling of surface RNAs across diverse mammalian blood cell types unveil abundant Y RNAs on the surface of human monocytes. Subsequent investigations uncover functional roles of surface Y RNAs on monocytes as "immune sentinel" to enrich extracellular histones, revealing a previously unknown pathway of innate immune activation. HighlightsNovel methodologies for cell surface RNA mapping, validation and functional interrogation. Sequencing of surface RNAs across various mammalian blood cell types offers detailed maps of surface RNAs. Abundant Y RNAs localize on human monocyte surfaces and capture extracellular histones released by adjacent ruptured cells. Surface Y RNAs enrich extracellular histones to facilitate transcription and secretion of IL-6 through NF-{kappa}B and AP-1 activation.

molecular biology↗

Identification of Novel Mutations Causing Clofazimine Resistance in Mycobacterium intracellulare

Clofazimine (CFZ) is a promising repurposed drug for treating Mycobacterium avium-intracellulare complex pulmonary disease (MAC-PD), but its resistance mechanisms in M. intracellulare remain poorly understood. In this study, we generated 36 CFZ-resistant M. intracellulare mutants in vitro and identified various mutations in the marR gene (WP_009952290.1) in 61% of resistant mutants through whole-genome sequencing. Mutations were identified in additional genes encoding flavin-dependent oxidoreductase (ssuD, C67A), membrane lipoprotein (lppI, C207 deletion), glucose-methanol-choline oxidoreductase (G157 deletion), MASE1 domain-containing protein (C62G), and PPE family protein (222C deletion). Gene complementation experiments demonstrated that introducing the wild-type marR in CFZ-resistant strains (L72 and L74) with marR mutations reduced CFZ minimum inhibitory concentrations (MICs) from 1 g/mL to the susceptible baseline (0.25 g/mL), confirming its critical role in CFZ resistance. Notably, the M. intracellulare MarR lacks homology to M. tuberculosis MarR family protein Rv0678 (MmpR) whose mutations are involved in CFZ and bedaquiline (BDQ) resistance and is flanked by non-efflux pump genes (dhmA and doxX), and unlike M. tuberculosis, its mutation does not cause bedaquiline cross-resistance, indicating a different MarR and distinct regulatory mechanism for CFZ resistance in M. intracellulare. This work highlights marR as a key determinant of CFZ resistance in M. intracellulare and underscores the need for further mechanistic studies with implications for clinical diagnostics and therapeutic strategies.

microbiology↗

Distinct yet Overlapping Functions of VMP1 and TMEM41B in Modulating Hepatic Lipoprotein Secretion and Autophagy in MASH

BackgroundTransmembrane protein 41B (TMEM41B) and vacuolar membrane protein 1 (VMP1) are endoplasmic reticulum (ER) scramblases that shuttle phospholipids between the inner and outer leaflets of the ER membrane. Both TMEM41B and VMP1 also play critical roles in regulating hepatic lipoprotein secretion and autophagy. Despite these similarities, whether TMEM41B and VMP1 exhibit different roles in very low-density lipoprotein (VLDL) secretion and autophagy in the pathogenesis of metabolic-associated steatotic liver disease (MASLD) remains unclear. MethodsWe created liver- and hepatocyte-specific single knockout (KO) and double knockout (DKO) mice for Tmem41b and Vmp1, as well as overexpression knock-in (KI) mice with hepatic overexpression of TMEM41B, Tmem41b KO/Vmp1 KI, and Vmp1 KO/Tmem41b KI. We conducted lipidomic, metabolomic, biochemical, and functional studies in these mice, fed either a chow diet or a MASLD diet. ResultsTMEM41B protein levels were decreased in the livers of human subjects with MASLD. The loss of hepatic Tmem41b impaired VLDL secretion, resulting in steatosis, inflammation, and fibrosis. Vmp1 KO mice exhibited similar phenotypes to DKO mice, displaying a more severe defect in VLDL secretion and greater liver injury than Tmem41b KO mice. Lipidomic analysis revealed decreased levels of phosphatidylcholine and phosphatidylethanolamine, along with increased neutral lipids in both Tmem41b KO and Vmp1 KO mice; however, these changes were generally more pronounced in Vmp1 KO mice. VMP1 and TMEM41B localized at the mitochondrial-associated membrane (MAM), and a reduction in mitochondria-ER contact was observed in hepatocytes deficient in either VMP1 or TMEM41B. Ultrastructural electron microscopy analysis showed increased accumulation of "lipid droplet" in the ER membrane bilayer and ER lumen in both Vmp1 KO and Tmem41b KO hepatocytes, with greater ER luminal "lipid droplet" accumulation in Tmem41b KO hepatocytes. The loss of hepatic Vmp1 or Tmem41b led to elevated levels of LC3-II and p62, with significantly higher levels of both markers in Vmp1 KO and DKO mouse livers compared to Tmem41b KO mouse livers. Restoring Vmp1 in Tmem41b KO mice partially improved defective VLDL secretion; however, high expression levels of VMP1 did not correct the hepatic autophagy defect. In contrast, restoring Tmem41b in Vmp1 KO mice dose-dependently enhanced both defective VLDL secretion and autophagy. Importantly, overexpression of hepatic TMEM41B mitigated diet-induced MASLD in mice. ConclusionThe loss of hepatic Vmp1 or Tmem41b decreases hepatic MAM and phospholipid content, leading to decreased VLDL secretion and promoting MASLD. While VMP1 and TMEM41B have overlapping functions, VMP1 appears to play a more critical role in regulating VLDL secretion and autophagy in mouse livers than TMEM41B.

cell biology↗

Deep Learning Predicts Survival Across Squamous Tumor Entities From H&E Stains: Insights from Head and Neck, Esophagus, Lung and Cervical Cancer

Computational pathology-based models are becoming increasingly popular for extracting biomarkers from images of cancer tissue. However, their validity is often only demonstrated on a single unseen validation cohort, limiting insights into their generalizability and posing challenges for explainability. In this study, we developed models to predict overall survival using haematoxylin and eosin (H&E) slides from FFPE samples in head and neck squamous cell carcinoma (HNSCC). By validating our models across diverse squamous tumor entities, including, head and neck (HR = 1.58, 95% CI = 1.17-2.12, p = 0.003), esophageal (non-significant), lung (HR = 1.31, 95% CI = 1.13-1.52, p < 0.001) and cervical (HR = 1.39, 95% CI = 1.10-1.75, p = 0.005) squamous cell carcinomas, we showed that the predicted risk score captures relevant information for survival beyond HNSCC. Correlation analysis indicated that the predicted risk score is strongly associated with various clinical factors, including human papillomavirus status, tumor volume and smoking history, although the specific factors vary across cohorts. These results emphasize the necessity for comprehensive validation and in-depth assessment of computational pathology-based models to better characterize the underlying patterns they learn during training.

pathology↗

BABA: Naturalistic fMRI and MEG recordings during viewing of a reality TV show

Prior neuroimaging datasets using naturalistic listening paradigms have predominantly focused on single-talker scenarios. While these studies have been invaluable for advancing our understanding of speech and language processing in the brain, they do not capture the complexities of real-world multi-talker environments. Here, we introduce the "Le Petit Prince (LPP) Multi-talker Dataset", a high-quality, naturalistic neuroimaging dataset featuring 40 minutes of electroencephalogram (EEG) and 7T functional magnetic resonance imaging (fMRI) recordings from 26 native Mandarin Chinese speakers as they listened to both single-talker and multi-talker speech streams. Validation analyses conducted on both EEG and fMRI data demonstrate the datasets high quality and robustness. Additionally, the dataset includes detailed transcriptions and prosodic and linguistic annotations of the speech stimuli, enabling fine-grained analyses of neural responses to specific linguistic and acoustic features. The LPP Multi-talker Dataset is well-suited for addressing a wide range of research questions in cognitive neuroscience, including selective attention, auditory stream segregation, and working memory processes in naturalistic listening contexts.

neuroscience↗

Ancestral proteins trace the emergence of substrate specificity and oligomerization within bacterial DEDDy dinucleases

Nucleases play a crucial role in bacterial physiology, influencing processes such as DNA repair, genome maintenance, and host-pathogen interactions. We recently identified a class of nucleases, diDNases, which are encoded on mobile genetic elements and homologous to the house-keeping nanoRNase C (NrnC). Despite their shared structural fold, diDNases and NrnC orthologs exhibit differences. DiDNases form dimers and preferably cleave DNA dinucleotides, whereas NrnC homologs assemble into octamers that do not discriminate between RNA or DNA dinucleotides. Here, we investigate the evolutionary divergence of these enzymes using ancestral sequence reconstruction. Our results show that both diDNases and NrnC orthologs originated from a dimeric ancestor with intermediate substrate preferences. Structural analyses of ancestral and extant dinucleases provide a molecular rational for how gradual changes in conformation gave rise to substrate preferences, oligomeric state, and catalytic efficiency of these related, yet distinct enzyme clades. These findings provide insights into how small structural modifications enable large-scale changes in molecular assembly and functional specialization harnessing a conserved protein fold. In addition, the preference of the early ancestors for DNA dinucleotides and preservation of this activity in all extant enzymes strongly argues for a biological function of DNA dinucleotides.

evolutionary biology↗

The Hitchdock domain in Kinesin-2 Tail Enables Adaptor Assembly and Cargo Binding

Intracellular transport relies on motor proteins like kinesins to deliver essential cargoes along microtubules, yet the mechanisms of cargo recognition remain unclear. Here, we present high-resolution cryo-electron microscopy structures of the heterotrimeric kinesin-2 complex (KIF3A/KIF3B/KAP3) bound to the adenomatous polyposis coli (APC) cargo. Our findings reveal a previously uncharacterized KIF3 tail motif, termed the "Hitchdock domain," which plays a pivotal role in mediating interactions with both the KAP3 adaptor and the APC cargo. In this domain, the KIF3A helical regions facilitate specific cargo binding, while the {beta}-hairpin region and KIF3B provide structural support. Mutagenesis and molecular dynamics simulations confirm the domains functional importance. Interestingly, the Hitchdock/KAP3 structure suggests a conserved structural basis for cargo recognition across molecular motors, including kinesin-1 and dynein, which utilize similar hook-like architectures, highlighting the potential universality of this mechanism. Furthermore, our findings provide insights into kinesin-2 cargo specificity and offer a molecular framework for understanding related diseases.

cell biology↗

Hydrogen-Induced Calcium Influx via the TRPC4-TRPC4AP Axis

BackgroundCalcium ions (Ca{superscript 2}) serve as universal intracellular messengers regulating diverse physiological processes, while dysregulated Ca{superscript 2} homeostasis triggers cytotoxicity. Molecular hydrogen (H2) exhibits protective effects against oxidative stress-related pathologies, but its mechanism of action remains incompletely understood. Transient receptor potential canonical 4 (TRPC4) channels and their associated protein TRPC4AP are critical mediators of Ca{superscript 2} influx ( [Ca{superscript 2}]i), yet their role in H2-mediated calcium signaling is unexplored. This study investigates the molecular mechanism by which H2 modulates Ca{superscript 2} dynamics through the TRPC4-TRPC4AP axis, aiming to establish its therapeutic potential for calcium-related disorders. MethodsThe study employed heterogeneous cellular models (e.g., mesenchymal stem cells, neurons, fibroblasts) and in vivo two-photon calcium imaging in C57BL/6J mice. Techniques included CRISPR-Cas9 knockout, siRNA-mediated gene silencing, molecular docking (AlphaFold 3), and protein-protein interaction analysis. Calcium flux was quantified via fluorescence imaging, while mitochondrial integrity and cytoskeletal dynamics were assessed using JC-1 staining, ATPase activity assays, and live-cell imaging. Structural validation of TRPC4-TRPC4AP binding sites utilized mutagenesis and complementation experiments. ResultsH2 selectively enhanced extracellular Ca{superscript 2} influx via TRPC4-TRPC4AP, with no cytotoxicity or mitochondrial dysfunction observed. Key arginine residues (730Arg-731Arg) in the TRPC4 CIRB domain formed hydrogen-bond networks essential for channel activation. In vivo, H2 increased neuronal Ca{superscript 2} transient frequency and amplitude in the primary motor cortex. TRPC4AP knockout abolished H2-induced Ca{superscript 2} influx, while mutagenesis of 730Arg/731Arg disrupted channel activity. H2 also promoted cytoskeletal remodeling and cell motility, dependent on TRPC4AP-mediated Ca{superscript 2} signaling. ConclusionsThis study identifies H2 as a novel calcium agonist that activates the TRPC4-TRPC4AP axis to regulate extracellular Ca{superscript 2} influx. The 730Arg-731Arg motif in TRPC4 serves as a critical H2-sensitive site, enabling dynamic calcium homeostasis without overload. These findings provide a mechanistic basis for H2-based therapies targeting calcium dysregulation in neurodegenerative, inflammatory, and metabolic diseases, while highlighting TRPC4AP as a pivotal molecular switch for gasotransmitter signaling.

biochemistry↗

Resilience and vulnerabilities of tumor cells under purine shortage stress

Purine metabolism is a promising therapeutic target in cancer; however how cancer cells respond to purine shortage,particularly their adaptation and vulnerabilities, remains unclear. Using the recently developed purine shortage-inducing prodrug DRP-104 and genetic approaches, we investigated these responses in prostate, lung and glioma cancer models. We demonstrate that when de novo purine biosynthesis is compromised, cancer cells employ microtubules to assemble purinosomes, multi-protein complexes of de novo purine biosynthesis enzymes that enhance purine biosynthesis efficiency. While this process enables tumor cells to adapt to purine shortage stress, it also renders them more susceptible to the microtubule-stabilizing chemotherapeutic drug Docetaxel. Furthermore, we show that although cancer cells primarily rely on de novo purine biosynthesis, they also exploit Methylthioadenosine Phosphorylase (MTAP)-mediated purine salvage as a crucial alternative source of purine supply, especially under purine shortage stress. In support of this finding, combining DRP-104 with an MTAP inhibitor significantly enhances tumor suppression in prostate cancer (PCa) models in vivo. Finally, despite the resilience of the purine supply machinery, purine shortage-stressed tumor cells exhibit increased DNA damage and activation of the cGAS-STING pathway, which may contribute to impaired immunoevasion and provide a molecular basis of the previously observed DRP-104-induced anti-tumor immunity. Together, these findings reveal purinosome assembly and purine salvage as key mechanisms of cancer cell adaptation and resilience to purine shortage while identifying microtubules, MTAP, and immunoevasion deficits as therapeutic vulnerabilities.

cancer biology↗

Augmenting microbial phylogenomic signal with tailored marker gene sets

Phylogenetic marker genes are traditionally selected from a fixed collection of whole genomes evenly distributed across major microbial phyla, covering only a small fraction of gene families. And yet, most microbial diversity is found in metagenome-assembled genomes that are unevenly distributed and harbor gene families that do not fit the criteria of universal orthologous genes. To address these limitations, we systematically evaluate the phylogenetic signal of gene families annotated from KEGG and EggNOG functional databases for deep microbial phylogenomics. We show that markers selected from an expanded pool of gene families and tailored to the input genomes improve the accuracy of phylogenetic trees across simulated and real-world datasets of whole genomes and metagenome-assembled genomes. The improved accuracy of trees compared to previous markers persists even when metagenome-assembled genomes lack a fraction of open reading frames. The selected markers have functional annotations related to metabolism, cellular processes, and environmental information processing, in addition to replication, translation, and transcription. We introduce TMarSel, a software tool for automated, systematic, free-from-expert opinion, and tailored marker selection that provides flexibility in the number of markers and annotation databases while remaining robust against uneven taxon sampling and incomplete genomic data.

bioinformatics↗

Enzyme Co-Scientist: Harnessing Large Language Models for Enzyme Kinetic Data Extraction from Literature

The extraction of molecular annotations from scientific literature is critical for advancing data-driven research. However, traditional methods, which primarily rely on human curation, are labor-intensive and error-prone. Here, we present an LLM-based agentic workflow that enables automatic and efficient data extraction from literature with high accuracy. As a demonstration, our workflow successfully delivers a dataset containing over 91,000 enzyme kinetics entries from around 3,500 papers. It achieves an average F1 score above 0.9 on expert-annotated subsets of protein enzymes and can be extended to the ribozyme domain in fewer than 3 days at less than $90. This method opens up new avenues for accelerating the pace of scientific research.

biochemistry↗

Scalability of Random Forest in Myoelectric Control

ObjectiveMyoelectric control systems translate electromyographic (EMG) signals into control commands, enabling immersive human-robot interactions in the real world and the Metaverse. The variability of EMG due to various confounding factors leads to significant performance degradation. Such variability can be mitigated by training a highly generalisable but massively parameterized deep neural network, which can be effectively scaled using a vast dataset. We aim to find an alternative simple, explainable, efficient and parallelisable model, which can flexibly scale up with a larger dataset and scale down to reduce model size, will significantly facilitate the practical implementation of myoelectric control. ApproachIn this work, we discuss the scalability of a random forest (RF) for myoelectric control. We show how to scale an RF up and down during the process of pre-training, fine-tuning, and automatic self-calibration. The effects of diverse factors such as bootstrapping, decision tree editing (pre-training, pruning, grafting, appending), and the size of training data are systematically studied using EMG data from 106 participants including both low- and high-density electrodes. Main resultsWe examined several factors that affect the size and accuracy of the model. The best solution could reduce the size of RF models by{approx} 500 x, with the accuracy reduced by only 1.5%. Importantly, for the first time we report the unique merit of RF that with more EMG electrodes (higher input dimension), the RF model size would be reduced, contrasting all other models. SignificanceAll of these findings unlock the full potential of RF in real-world applications.

bioengineering↗