Biology subjects

Ma, S.

Publications and source records attributed to Ma, S..

At least 55 records · Page 3Linked to original sources

bioRxiv2026

Targeting insulo-frontal pathway to reduce stress-evoked cognitive rigidity

The anterior insular cortex (aIC), a central hub of the salience network, is engaged by cognitive flexibility tasks1,2; it is also implicated in stress-related mental disorders3-5, where cognitive rigidity is a common but poorly treated symptom1. While the insular cortexs roles in interoception and emotional regulation are extensively studied6-10, its causal contribution to cognitive rigidity remains unclear. Using attentional set-shifting tasks (AST) in mice, we identify aIC neurons projecting to the medial prefrontal cortex (mPFC) as key regulators of adaptive decision-making. These neurons show heightened activity following incorrect--but not correct--trials. This elevated activity persists into subsequent trials, providing a salience signal that enhances mPFC outcome-dependent updating and promotes convergence of neural activity patterns across trials. Optogenetic manipulation of aIC[->]mPFC projections during the pre-decision phase disrupts mPFC updating and impairs AST performance. Moreover, stress disrupts the outcome-dependence of aIC activity and impairs set-shifting. Crucially, selectively reinforcing aIC[->]mPFC activity after incorrect trials via optogenetics enhances mPFC updating, improves neural activity convergence across trials, and restores cognitive flexibility in stressed mice. These findings reveal a previously unrecognized role of the aIC[->]mPFC circuit in linking trial outcomes to adaptive decision-making and identify this pathway as a promising target for treating stress-induced cognitive rigidity.

neuroscience↗

bioRxiv2026

A CRISPR-Based Humanized Model Reveals Cooperative Role of STAG2 Loss in Familial GATA2-Deficient MDS Progression

Myelodysplastic syndrome (MDS) is a heterogeneous myeloid malignancy driven by hematopoietic stem cell dysfunction, leading to ineffective hematopoiesis and cytopenias. Familial GATA2 deficiency is the most common cause of Myelodysplastic syndrome in adolescents, with progression often accelerated by co-occurring mutations, notably STAG2 loss-of-function. Using CRISPR/Cas9-mediated genome engineering in primary human fetal liver-derived hematopoietic stem cells and xenotransplantation in mice, we modeled GATA2-deficient Myelodysplastic syndrome with acquired STAG2 loss to investigate disease initiation and progression. While GATA2 deficiency alone had minimal short-term impact in our model, combined GATA2 and STAG2 loss increased hematopoietic stem cell maintenance and self-renewal, induced a myeloid-lineage bias, and expanded primitive progenitors. Single-cell transcriptional profiling revealed upregulation of stemness genes and inflammatory pathways. This humanized model faithfully recapitulates high-risk GATA2-deficient Myelodysplastic syndrome, providing mechanistic insight into how cooperative mutations drive stem cell expansion, inflammatory signaling, and myeloid skewing. Visual Abstract O_FIG O_LINKSMALLFIG WIDTH=200 HEIGHT=119 SRC="FIGDIR/small/702879v1_ufig1.gif" ALT="Figure 1"> View larger version (20K): org.highwire.dtl.DTLVardef@1972beforg.highwire.dtl.DTLVardef@1c57886org.highwire.dtl.DTLVardef@16bd582org.highwire.dtl.DTLVardef@8e728a_HPS_FORMAT_FIGEXP M_FIG C_FIG Key PointsO_LIHumanized model of familial GATA2-deficiency requires the loss of STAG2 for progression to an MDS disease phenotype C_LIO_LIGATA2-ko+STAG2-ko increase HSC self-renewal, induce a myeloid-lineage bias, and trigger an inflammatory transcriptional program C_LI

cancer biology↗

bioRxiv2026

Early microglial priming in Alzheimer's disease revealed by ME-seq

Epigenetic modifications, particularly DNA methylation, change dynamically with aging and are implicated in Alzheimers Disease (AD), yet how methylation interfaces with transcriptional and chromatin regulation at single-cell resolution remains poorly understood. Progress has been limited by a lack of scalable technologies capable of jointly profiling these regulatory layers. Here, we present ME-seq, a highly scalable technologies capable of simultaneously profiling DNA methylation, gene expression, and chromatin accessibility, while achieving a 100-fold reduction in cost. We generated over 400,000 single-nucleus trimodal profiles from the aging and AD mouse brain across ages, producing the first such atlas of neurodegeneration. We found AD progression triggers pronounced, disease-specific shifts in cellular composition, characterized by accelerated epigenetic aging and the expansion of disease-associated microglia (DAM). Integrative analyses, including aging clocks, revealed that DNA methylation acts as an early priming layer preceding transcriptional activation with IRF1 identified as a methylation-sensitive transcription factor serving as a gatekeeper for DAM activation. Our results establish ME-seq as a transformative tool for large-scale epigenomic dissection, revealing DNA methylation as a primary coordinator of cell-state transitions in the aging brain. Graphic abstract O_FIG O_LINKSMALLFIG WIDTH=195 HEIGHT=200 SRC="FIGDIR/small/702946v1_ufig1.gif" ALT="Figure 1"> View larger version (70K): org.highwire.dtl.DTLVardef@67435dorg.highwire.dtl.DTLVardef@177c22aorg.highwire.dtl.DTLVardef@16c6fd9org.highwire.dtl.DTLVardef@231b92_HPS_FORMAT_FIGEXP M_FIG C_FIG

genomics↗

bioRxiv2026

BiHiTo: Biomolecular Hierarchy-inspired Tokenization

Three-dimensional atomic arrangements of biomolecules are key to demystifying biological functions. The rapid expansion of accessible structural data, driven by advances in AI for science, highlights the critical challenge of efficiently modeling large-scale biomolecular structures, which are high-dimensional systems shaped by biological assembly principles. To address this, we introduce BiHiTo, a multi-level Biomolecular Hierarchy-inspired Tokenizer that intrinsically mimics natural biological assembly hierarchies. Specifically, we design a multi-codebook quantizer that mirrors the natural hierarchy of biomolecular structure, enabling simultaneous capture of representations spanning atomic motifs to global conformational variations. This hierarchical alignment markedly improves the biological interpretability and reconstruction fidelity of biomolecular structure.Extensive experiments demonstrate that BiHiTo delivers state-of-the-art performance and robust generalization across molecular dynamics trajectories and macromolecular complexes, facilitating advances in structure generation and dynamic conformation exploration. In the reconstruction of the CASP14 and OOD test set FastFolding protein multi-conformation data, our method achieves a 17% and 51% reduction in RMSD compared to Bio2Token, respectively.

molecular biology↗

bioRxiv2026

Organoid Profiler: Automated, high-throughput and quantitative morphological characterization uncovers conserved longitudinal developmental kinetics in microfluidics-engineered organoids

Three-dimensional organoid systems have emerged as transformative tools in biomedical research, yet their translation into standardized high-throughput screening platforms is impeded by stochastic fabrication variability and a lack of scalable quantitative analytics. Here, we present a unified framework integrating microfluidic droplet engineering with Organoid Profiler, an automated deep-phenotyping pipeline designed to capture longitudinal morphological kinetics. By processing over 10,000 longitudinally tracked images across murine lung and liver and human cerebral models, we show that this platform achieves human-level segmentation precision (r = 0.99) while uncovering conserved developmental dynamics obscured in manual cultures. We identify a distinct biphasic "remodeling-to-expansion" trajectory in microfluidic organoids, characterized by an initial compaction phase underpinned by the transcriptional upregulation of focal adhesion pathways, followed by exponential expansion. We further demonstrate a "shape relaxation" phenomenon where tissues converge to spherical equilibrium regardless of initial geometry, indicating intrinsic self-organization. Multi-modal profiling confirms that microfluidic organoids exhibit superior viability, homogeneity (area variability [≤] 5% upon fabrication, [≤] 15% across development) and biological fidelity, retaining physiological levels of immune and stromal niche components often lost in conventional culture. This study establishes morphological metrics as reliable, non-invasive proxies for tissue health and molecular maturity, providing a standardized foundation for quantitative organoid biology.

bioengineering↗

bioRxiv2025

Liver Zonation Disruption Fuels Hepatocellular Carcinoma in Chronic Liver Disease

Hepatocellular carcinoma (HCC) arises almost exclusively in chronic liver disease (CLD), yet the classical etiological drivers of injury insufficiently explain why only a subset of patients progress to cancer. Here, we identify disruption of liver metabolic zonation, specifically, aberrant expansion of {beta}-catenin activity from pericentral to periportal territories as a previously unrecognized tumorigenic risk state that emerges across etiologically diverse CLDs. Using spatial transcriptomics and immunohistochemistry from murine liver disease followed by human sample validation, we demonstrate that MASLD/MASH, alcohol-associated hepatitis, viral hepatitis, and immune-mediated cholangiopathies share a striking periportal induction of pericentral {beta}-catenin target programs, indicating a conserved zonation disturbance independent of disease etiology. Since {beta}-catenin expansion occurs alongside inflammation, fibrosis, and metabolic dysfunction in human CLD, its direct oncogenic contribution remained unclear. To functionally isolate the tumorigenic consequence of zonation disruption itself from these confounding disease processes, we employed hepatocyte-specific deletion of ZNRF3 and RNF43--negative regulators that physiologically restrict {beta}-catenin to the pericentral zone. Lineage tracing demonstrates that ZNRF3/RNF43 deletion drives selective periportal hepatocyte proliferation, zonal reprogramming, and tumor initiation in a {beta}-catenin-dependent manner, establishing zonation disruption as a direct mechanistic driver of carcinogenesis. The resulting tumors exhibit a distinct metabolic and immunologic phenotype, including heightened mitochondrial respiration, preserved periportal identity, and T-cell competence, and correspond to a molecularly defined subset comprising 5-10% of human HCCs. Together, these findings reveal {beta}-catenin zonation expansion as a conserved and previously unrecognized HCC risk factor that mechanistically links chronic liver injury to malignant transformation. They further establish ZNRF3/RNF43 deletion as a tractable model of zonation-driven hepatocarcinogenesis and identify a distinct human HCC subtype with unique therapeutic vulnerabilities, opening new avenues for mechanism-based risk stratification, early detection and preventive therapeutic strategies in patients with chronic liver disease.

cancer biology↗

bioRxiv2025

BASSL-MI: Batch-Agnostic Self-Supervised Learning Uncovers Clinically Relevant Tumor Niches in Multiplexed Imaging

Multiplexed imaging enables rich, high-resolution characterization of the tumor microenvironment but relies on labor-intensive and error-prone cell segmentation and phenotyping pipelines. We present BASSL-MI, a batch-agnostic, self-supervised framework for discovering tissue niches directly from multiplexed imaging data. BASSL-MI operates directly on image patches, eliminating the need for explicit cell segmentation while mitigating image-, sample-, or batch-specific artifacts. Built on a modified contrastive block disentanglement architecture, BASSL-MI learns dual latent representations that separate biologically informative features from batch-dependent factors through spatially guided augmentations and batch-invariance objectives. Applied to a 56-marker colorectal cancer CODEX dataset, BASSL-MI-trained embeddings markedly reduce image-specific variability and recover biologically interpretable spatial niches. Notably, it uncovers CD20-rich follicular regions associated with improved survival, outperforming published findings from cell segmentation-driven clustering. This work demonstrates that self-supervised, patch-based learning can capture clinically relevant spatial organization within tumor microenvironments, advancing toward automated, non-cell-based analysis of multiplexed imaging data.

bioinformatics↗

bioRxiv2025

Argon inhibits autophagy and improves cerebral ischemia-reperfusion injury via PI3K/Akt/mTOR pathway

It has been reported that argon has neuroprotective effects during cerebral ischemia-reperfusion injury (CIRI). However, the specific mechanism is not fully understood. The aim of this study is to investigate if argon could regulate autophagy to exert a neuroprotective function. A temporary middle cerebral artery ischemia mouse model (tMCAO, ischemia 3h, reperfusion 1h) and the HT22 cell line Oxygen glucose deprivation reperfusion (OGDR) model were used for argon treatment. Transcriptome sequencing and bioinformatics analysis were performed on HT22 OGDR models. Western blot, CCK-8, LDH, transmission electron microscopy, agonists and inhibitors of autophagy and PI3K/Akt/mTOR were used to investigate the specific mechanism. Argon could reduce infarct volume, improve weight recovery, and neurological scores. Enrichment analysis of the KEGG signaling pathway was highly correlated with the PI3K/Akt pathway. Western blot, CCK-8, LDH, and transmission electron microscopy showed that argon could activate the PI3K/Akt/mTOR pathway and inhibit autophagy, promoting a neuroprotective function. After the combined use of autophagy inducer 3, Akt inhibitor MK2206, and mTOR inhibitor rapamycin, respectively, the neuroprotective function of argon was significantly reversed. Our study suggests that argon activating the PI3K/Akt/mTOR pathway and inhibiting autophagy is an important mechanism for its neuroprotective function.

neuroscience↗

bioRxiv2025

A gene program dictionary of human cells

Defining all human cell types and their roles in health and disease is a central goal of biology. Single-cell RNA sequencing has enabled the construction of organ-specific cell atlases, but building a comprehensive organism-wide atlas spanning multiple organs remains challenging due to batch effects, study biases, and inter-organ complexity. Here, we present Gene Program Dictionary (GPD), a framework that leverages robust gene co-expression programs--rather than direct cell integration--to overcome these barriers. Using SpacGPA, a partial correlation-based network method, we analyzed 466 scRNA-seq datasets, generating 1,975 independent networks and 90,701 gene co-expression modules, which were consolidated into 1,534 consensus gene programs representing a wide range of human tissues and cell types. Each program serves as a composite marker, capturing both cell-type-specific and shared biological processes. We demonstrate their utility by mapping endothelial cell subtypes across tissues to reveal their heterogeneity--including tumor-specific programs--annotating colorectal cancer spatial transcriptomes, and linking programs and their corresponding cell types to disease loci, revealing hotspots such as neuronal programs in psychiatric disorders and a proximal tubule program in kidney diseases. GPD provides an organism-wide reference for studying cellular diversity and disease mechanisms.

bioinformatics↗

bioRxiv2025

Automatic Classification of Circulating Blood Cell Clusters based on Multi-channel Flow Cytometry Imaging

Circulating blood cell clusters (CCCs) containing red blood cells (RBCs), white blood cells (WBCs), and platelets are significant biomarkers linked to conditions like thrombosis, infection, and inflammation. Flow cytometry, paired with fluorescence staining, is commonly used to analyze these cell clusters, revealing cell morphology and protein profiles. While computational approaches based on machine learning have advanced the automatic analysis of single-cell flow cytometry images, there is a lack of effort to build tools to automatically analyze images containing CCCs. Unlike single cells, cell clusters often exhibit irregular shapes and sizes. In addition, these cell clusters often consist of heterogeneous cell types, which require multi-channel staining to identify the specific cell types within the clusters. This study introduces a new computational framework for analyzing CCC images and identifying cell types within clusters. Our framework uses a two-step analysis strategy. First, it categorizes images into cell cluster and non-cluster groups by fine-tuning the You Only Look Once(YOLOv11) model, which outperforms traditional convolutional neural networks (CNNs), Vision Transformers (ViT). Then, it identifies cell types by overlaying cluster contours with regions from multi-channel fluorescence stains, enhancing accuracy despite cell debris and staining artifacts. This approach achieved over 95% accuracy in both cluster classification and phenotype identification. In summary, our automated framework effectively analyzes CCC images from flow cytometry, leveraging both bright-field and fluorescence data. Initially tested on blood cells, it holds potential for broader applications, such as analyzing immune and tumor cell clusters, supporting cellular research across various diseases.

bioinformatics↗

bioRxiv2025

Interface-Sensitive Epi-Coherent Anti-Stokes Raman Scattering Microscopy for Imaging Cell Adhesion Dynamics

Studying cell adhesion dynamics is critical to understand how cells interact with the extracellular matrix and how they migrate. Conventional microscopy techniques for interfacial studies either lack chemical information or require fluorescence labeling. We found that homodyne epi-coherent anti-Stokes Raman scattering (epi-CARS) microscopy, when focused as interfaces, has intrinsic interfacial selectivity. The interference of the back-reflected nonresonant signals amplifies the contrast of cell adhesion areas, generating a distinctive negative contrast at attachment areas, which reflects the distance between the cell membrane and the substrate with tens of nanometers precision. The Epi-CARS configuration also provides high contrasts for cellular lipid droplets when applied in the inverted microscope configuration. Furthermore, by incorporating a pinhole into our confocal epi-CARS system, we can effectively reject out-of-focus reflections that degrade epi-CARS image quality, allowing us to amplify adhesion and lipid droplet contrasts for complex samples, such as cells. We applied this method to study cell-substrate adhesion dynamics of mitotic cells, revealing adhesion site splitting during mitosis and their outward development during postmitotic spreading. Furthermore, we found that after mitosis, the leading edges exhibited homogeneous adhesion regions with thicker water layers, whereas the retracting edges showed heterogeneous adhesion patterns with thinner layers at the contact sites.

biochemistry↗

bioRxiv2025

SUGP1 associates Y-box protein to regulate piRNA biogenesis in Bombyx mori

PIWI-interacting RNAs (piRNAs) are critical for transposon silencing and genome integrity, as well as gene expression regulation and antiviral immunity in metazoans, yet the molecular mechanisms governing their biogenesis remain incompletely understood. The participation of the splicing-associated process in piRNA biogenesis has been emphasized in multiple species, but the key factors and mechanisms remain elusive. Here, we identified SUGP1 (SURP and G-Patch Domain Containing 1) in BmE (a unique model cell system with a complete piRNA biogenesis pathway) as a key splicing factor that functions in piRNA biogenesis. Through CRISPR-Cas9-mediated gene knockdown in cultured cells combined with RNA-seq, small RNA-seq, and IP-mass spectrometry (IP-MS), our results reveal that SUGP1 deficiency disrupts piRNA accumulation, alters mature piRNA length distributions, and activates transposon expression. Immunofluorescence and Western blot (WB) analyses further demonstrate that SUGP1 interacts with Y-box protein (YBP), which is key regulators of RNA metabolism. Functional validation in Drosophila SUGP1-RNAi lines highlights evolutionary conserved and species-specific roles of SUGP1 in piRNA maturation. Collectively, our data uncover a dual role for silkworm SUGP1 in coordinating YBP-dependent piRNA biogenesis, thus elucidating a novel mechanistic framework for piRNA pathway regulation. Our work also underscores the silkworm as a unique model for studying non-canonical piRNA biogenesis mechanisms, with implications for treating transposon dysregulation-linked diseases. Author summaryDisruption of piRNA synthesis leads to abnormal consequences such as transposon de-repression, posing a significant threat to genomic stability. It is therefore essential to in-depth analysis of the piRNA biosynthetic mechanism. Previous studies have highlighted the involvement of splicing-related processes in piRNA biosynthesis, yet key factors and mechanisms remain poorly understood. This study employed the silkworm cell system, which possesses a complete piRNA biosynthetic pathway. Utilizing CRISPR-Cas9-mediated gene knockout technology combined with RNA sequencing, small RNA sequencing, and immunoprecipitation-mass spectrometry (IP-MS), we discovered that SUGP1 deficiency disrupts piRNA accumulation, alters the length distribution of mature piRNAs, and activates transposon expression. Furthermore, immunofluorescence and Western blot analyses confirmed an interaction between SUGP1 and Y-box proteins (YBPs), key regulators of RNA metabolism. Functional validation using Drosophila SUGP1-RNAi lines revealed that SUGP1 plays dual roles in piRNA maturation. Our study reveals that the silkworm scissor-related factor SUGP1 has dual functions in coordinating YBP-dependent piRNA biosynthesis.

molecular biology↗

bioRxiv2025

SpacGPA: annotating spatial transcriptomes through de novo interpretable gene programs

Spatial transcriptomes, especially those from high-density platforms (e.g., Visium HD, Stereo-seq, Xenium Prime 5K), enable detailed tissue mapping but remain challenging to annotate for spatial domains and spatially variable genes (SVGs). Most existing methods depend on companion single-cell datasets and are sensitive to cell segmentation. We present SpacGPA, a segmentation-free, single-cell-independent framework that identifies de novo interpretable gene programs--co-expressed gene modules capturing key biological processes--from spatial transcriptomes. Leveraging GPU-acceleration and block-matrix operations, SpacGPA efficiently detects gene programs based on a graphical Gaussian model in ultra-large datasets. These programs exhibit domain-specific spatial expression patterns and cell type-specific expression in independent single-cell datasets, serving as robust markers for spatial domains. Applied to four mouse and human spatial datasets, SpacGPA identified 58-109 gene programs per dataset, revealing refined tissue structures. Hub genes within these programs correlate with canonical cell type markers and SVGs, enabling systematic SVG identification. The programs are also functionally enriched and transferable across datasets and platforms. Built on AnnData, SpacGPA provides a scalable, low-hardware solution for high-resolution annotation of tissue organization and gene regulation.

bioinformatics↗

bioRxiv2025

A single-cell spatial transcriptomic census of human skin anatomy

The skin is the largest human organ and a site of significant disease burden, yet its cellular and molecular organization across the body are largely undefined. Here, we construct a spatially-resolved single-cell atlas of 1.2 million cells from normal adult human skin to localize 45 cell types across 15 anatomic sites. We define principles of organ-wide cell composition, including axes of cell diversity and specialization, and distinguish site-enriched cell types. Each body site is comprised of 10 multicellular neighborhoods that define cell-cell communication. Notably, we identify a perivascular neighborhood enriched for immune-stromal crosstalk with features resembling a homeostatic immune niche similar to skin-associated lymphoid tissue. Finally, mapping these neighborhoods onto skin disease reveals pathogenic neighborhood disruptions, including pan-disease immune alterations in the perivascular neighborhood. We present a framework charting the skins multiscale spatial organization across a molecular to macroanatomic scale. This work advances our understanding of organ-wide skin cellular organization and communication, and its architectural disruption in disease. HIGHLIGHTSO_LIHuman skin MERFISH spatial atlas of 1.2 million cells from 114 samples and 22 donors C_LIO_LICellular diversity varies along a central to peripheral body site spatial axis C_LIO_LI10 multicellular neighborhoods define skin microanatomy and homeostatic interactions C_LIO_LIHuman skin diseases feature spatial and transcriptional neighborhood remodeling C_LI

genomics↗

bioRxiv2025

Gliomas phenocopy an inborn error of metabolism to drive neuronal activity and tumor growth

The metabolic hallmarks of high-grade glioma (HGG) are not fully understood. Human brain tissue metabolomics revealed that the creatine synthesis pathway intermediate guanidinoacetate (GAA) accumulated [~]100-fold in HGGs relative to controls, which was caused by imbalanced activities of enzymes in this pathway. Glioma cells secreted GAA rather than using it to produce creatine, implicating an unexpected function. GAA accumulates in GAMT deficiency, an inborn error of metabolism, and elevates neuronal excitability. Neuronal excitability is also increased in glioma and drives tumor growth through neuron-glioma interactions. We hypothesized that glioma-generated GAA excites surrounding neurons. Indeed, GAA induced neuronal hyperactivity by activating GABAA receptors and causing depolarizing GABA currents in glioma-associated neurons with dysregulated chloride homeostasis. Depleting tumoral GAA decreased electrochemical activity, neuron-glioma interactions, and tumor aggressiveness. Our findings unveil a new mechanism linking cancer metabolism with cancer neuroscience and leverage human genetics to nominate GAA synthesis as a target in gliomas.

cancer biology↗

bioRxiv2025

Human-specific features of the cerebellum and ZP2-regulated synapse development

Understanding the unique features of the human brain compared to non-human primates has long intrigued humankind. The cerebellum refines motor coordination and cognitive functions, contributing to the evolutionary development of human adaptability and dexterity. To identify shared and divergent features across primates, we conducted single-nucleus transcriptomic and chromatin accessibility profiling of the adult cerebellar cortex in humans, chimpanzees, macaques, and marmosets. We revealed human-specific transcriptomic and regulatory features, particularly those involved in synaptogenesis. Notably, we identified an enrichment of the sperm receptor zona pellucida glycoprotein 2 (ZP2) and its potential interactors, known for their roles in gamete interaction, in human granule cells. Experimental data show that ZP2 expression in human granule cells is induced by pontine mossy fibers, reducing synaptic proteins at pontocerebellar glomerular synapses, and decreasing cerebellar neuron electrophysiological activity. This unexpected co-option of ZP2 in human-specific synapse regulation provides insights into the evolutionary specialization of the human cerebellum.

neuroscience↗

bioRxiv2025

Empirical optimization of dual-sgRNA design for in vivo CRISPR/Cas9-mediated exon deletion in mice

CRISPR/Cas9 has transformed gene editing, enabling precise genetic modifications across species. However, existing sgRNA design prediction models based on in vitro data are difficult to generalize to in vivo contexts. In particular, approaches based on single-sgRNA design require additional filtering of in-frame mutations, which is inefficient in terms of both time and cost. In this study, we developed the first mammalian in vivo-trained prediction model to evaluate the efficiency of a dual-sgRNA-based exon deletion strategy. Using 230 editing outcomes of postnatal viable individuals, eight prediction models were constructed and evaluated based on generalized linear models and random forests. The final selected model, a Combined GLM, integrated the DeepSpCas9 score with k-mer sequence features, achieving an AUC of 0.759 (95% Confidence Interval: 0.697-0.821). Motif analysis revealed that CC sequences were associated with high efficiency and TT sequences were associated with low editing efficiency. This study demonstrates that integrating sequence-based features with existing design scores can improve sgRNA efficiency prediction in vivo. The proposed framework can be applied to the development of next-generation sgRNA design tools, with direct implications for gene therapy, effective animal model generation, and precision genome engineering. GRAPHICAL ABSTRACT O_FIG O_LINKSMALLFIG WIDTH=200 HEIGHT=100 SRC="FIGDIR/small/675005v1_ufig1.gif" ALT="Figure 1"> View larger version (21K): org.highwire.dtl.DTLVardef@12e6170org.highwire.dtl.DTLVardef@1ff2599org.highwire.dtl.DTLVardef@1fd61dborg.highwire.dtl.DTLVardef@23a426_HPS_FORMAT_FIGEXP M_FIG C_FIG

bioinformatics↗

bioRxiv2025

A Machine Learning Approach to Predicting Dyspnea with Noninvasive Biomarkers

RationaleDyspnea is the subjective sensation of breathing discomfort. This symptom is highly prevalent in patients with chronic and critical illness, and its presence is associated with poor clinical outcomes and long-term psychological trauma. The multidimensional nature of the neurophysiological mechanisms underlying dyspnea, paired with individual variation in its presentation, makes identifying and monitoring this symptom difficult, particularly in non-communicative patients. ObjectiveUndetected and untreated dyspnea in critically ill patients is a significant problem contributing to patient suffering. Therefore, the objective of this study was to investigate the feasibility of machine learning methods for assessing and continuously monitoring dyspnea using easily obtained noninvasive biomarkers. MethodsWe recruited healthy participants (N = 60, 35 women) and stimulated dyspnea using a forced end-tidal semi-rebreathing circuit to modulate arterial oxygen and carbon dioxide levels while collecting non-invasive biomarker data and continuous self-reported dyspnea severity scores. This data was used to train machine-learning models to predict the presence or absence of significant dyspnea (Numeric Rating Scale [≥] 3). We then compared the performance of our final model to observational estimates by trained healthcare providers. Measurements and Main ResultsThe final model (Random Forest) performed well (PR-AUC=0.832) and exceeded the accuracy of observation estimates made on the same participants using the Respiratory Distress Observational Scale (RDOS) (accuracy=54%). ConclusionsThese results indicate that machine learning models can utilize non-invasive biomarker inputs to accurately predict carbon dioxide- and hypoxia-induced dyspnea in a healthy population during spontaneous breathing.

physiology↗