Biology subjects

Huang, Y.-A.

Publications and source records attributed to Huang, Y.-A..

8 recordsLinked to original sources

bioRxiv2026

CSGDA: A Cell State-Guided Graph Domain Adaptation Network for Single-Cell Drug Response Prediction

Intratumoral heterogeneity drives cancer recurrence and metastasis, yet single-cell drug response prediction faces severe "cross-domain" challenges, such as applying in vitro models to in vivo tissues or inferring metastatic resistance from primary tumors. These scenarios trigger distribution shifts arising from heterogeneous sequencing platforms, distinct tissue microenvironments, and metastatic evolution--problems rarely addressed by existing methods. We introduce CSGDA, a cell state-guided graph domain adaptation framework designed to predict drug responses across these biological heterogeneities. CSGDA incorporates biological priors to map gene expression into functional cell states, guiding a structure learning module to construct robust cell topology. To conquer distribution shifts, the model employs graph domain adaptation combined with a novel overlap penalty mechanism. Extensive benchmarks on five scRNA-seq datasets demonstrate that CSGDA outperforms state-of-the-art methods, achieving an average gain of [~]6% in ACC and AUPR. Beyond prediction accuracy, we employed integrated gradients to effectively pinpoint key genes involved in drug resistance within a challenging cross-metastasis cisplatin dataset. These findings underscore CSGDAs superior performance in single-cell drug response prediction and its potential in resolving single-cell heterogeneity, paving the way for precision medicine.

bioinformatics↗

bioRxiv2025

OfUSA: OpenfUS Analyzer, a versatile open-source framework for the analysis and visualization of functional ultrasound imaging data across animal models

Functional ultrasound (fUS) imaging is rapidly gaining interest for its unprecedented ability to study large-scale brain dynamics, yet its adoption and broader dissemination have been hindered by the lack of standardized tools and methodologies to analyze and interpret its rich datasets. We present OpenfUS Analyzer (OfUSA), a companion software suite designed to help researchers quickly engage with fUS data and perform the full range of analyses needed to generate publication-ready results and figures without relying on additional software. OfUSA offers a versatile and modular architecture including preprocessing, recording quality assessment, signal dynamics exploration, statistical analysis and visualization. These functions are separated yet easily combined into analytic pipelines through a programming-free graphical interface. The framework can be applied across species and experimental contexts, either by registering data to anatomical atlases, as shown here for the mouse brain, or by analyzing data without atlas constraints, as illustrated in a primate dataset. This flexibility, together with its comprehensive functionality, makes OfUSA a practical solution for standardized and reproducible analysis of fUS data in both preclinical and translational research. Using OfUSA, we demonstrate the capacity to detect stimulus-evoked responses with high sensitivity, identify their spatial localization within brain networks, and quantify both their extent and temporal dynamics. These results highlight the softwares ability to capture robust activation patterns and provide detailed insights into brain function, thereby accelerating the use of fUS as a powerful tool for systems neuroscience. HighlightsWe present OpenfUS Analyzer (OfUSA), a novel software platform for the complete analysis of functional ultrasound (fUS) datasets. OfUSA combines a user-friendly graphical interface with a standardized, flexible workflow and powerful visualization tools, making it an ideal solution for fUS researchers at all experience levels. The softwares utility is first demonstrated through the analysis of rodent fUS data using a standardized atlas, while its versatility is further emphasized by the successful analysis of a primate fUS dataset without a template, thereby illustrating its adaptability to non-standard experimental conditions.

neuroscience↗

bioRxiv2025

Activation of IL-17+ ILC subsets in IL-18R-deficient mice during fungal allergen exposure

Group 2 innate lymphoid cells (ILC2s) are critical players during type 2 inflammation present in most forms of asthma. ILC2s are tissue-resident cells that produce cytokines IL-5 and IL-13 critical to eosinophilic airway inflammation, mucus production, remodeling, and hyperresponsiveness. Though each ILC subset (ILC1s, ILC2s, ILC3s) is identified by specific transcription factors, cell surface receptors and cytokine profiles, functional plasticity between ILC subtypes occurs in various contexts. IL-18/IL-18R loci SNPs are linked to asthma in multiple genome-wide association studies and IL-18 has been shown to promote plasticity in ILC2s. Despite this, little is known about the in vivo role of IL-18/IL-18R on ILC2 responses in the lung. Within hours after mice were exposed to the fungal allergy Alternaria alternata, airway levels of IL-18 and IL-18 receptor expression increased on ST2+ ILCs. Single-cell RNA sequencing of lung cells from Alternaria-challenged mice showed that Il18 was largely expressed by alveolar macrophages, while IL-18R was highly expressed in IL-13+ILC2s. Utilizing IL-18 receptor knock-out mice (IL- 18R-/-), we observed increases in IL-17A production from both ST2+ and ST2-negative ILCs compared to control mice. We further observed an early increase in dual production of IL-5 and IL-17A in ST2+ ILCs followed by enhanced lung eosinophilia in the absence of IL-18R. Together, our findings suggest that IL-18 signaling prevents IL-17A production from ILC2s and subsequent eosinophilia in vivo. A further understanding of the regulation of ILC plasticity may lead to novel therapeutic targets in the treatment of ILC-driven asthma.

immunology↗

bioRxiv2025

scKAN: Interpretable Single-cell Analysis for Cell-type-specific Gene Discovery and Drug Repurposing via Kolmogorov-Arnold Networks

Single-cell analysis has revolutionized our understanding of cellular heterogeneity, yet current approaches face challenges in efficiency and interpretability. In this study, we present scKAN, a framework that leverages Kolmogorov-Arnold Networks for interpretable single-cell analysis through three key innovations: efficient knowledge transfer from large language models through a lightweight distillation strategy; systematic identification of cell-type-specific functional gene sets through KANs learned activation curves; and precise marker gene discovery enabled by KANs importance scores with potential for drug repurposing applications. The model achieves superior performance on cell-type annotation with a 6.63% improvement in macro F1 score compared to state-of-the-art methods. Furthermore, scKANs learned activation curves and importance scores provide interpretable insights into cell-type-specific gene patterns, facilitating both gene set identification and marker gene discovery. We demonstrate the practical utility of scKAN through a case study on pancreatic ductal adenocarcinoma, where it successfully identified novel therapeutic targets and potential drug candidates, including Doconexent as a promising repurposing candidate. Molecular dynamics simulations further validated the stability of the predicted drug-target complexes. Our approach offers a comprehensive framework for bridging single-cell analysis with drug discovery, accelerating the translation of single-cell insights into therapeutic applications.

bioinformatics↗

bioRxiv2024

PIP-Seq identifies novel heterogeneous lung innate lymphocyte population activation after combustion product exposure

Innate lymphoid cells (ILCs) are a heterogeneous population that play diverse roles in airway inflammation after exposure to allergens and infections. However, how ILCs respond after exposure to environmental toxins is not well understood. Here we show a novel method for studying the heterogeneity of rare lung ILC populations by magnetic enrichment for lung ILCs followed by particle-templated instant partition sequencing (PIP-seq). Using this method, we were able to identify novel group 1 and group 2 ILC subsets that exist after exposure to both fungal allergen and burn pit-related constituents (BPC) that include dioxin, aromatic hydrocarbon, and particulate matter. Toxin exposure in combination with fungal allergen induced activation of specific ILC1/NK and ILC2 populations as well as promoted neutrophilic lung inflammation. Oxidative stress pathways and downregulation of specific ribosomal protein genes (Rpl41 and Rps19) implicated in anti-inflammatory responses were present after BPC exposure. Increased IFN{gamma} expression and other pro-neutrophilic mediator transcripts were increased in BPC-stimulated lung innate lymphoid cells. Further, the addition of BPC induced Hspa8 (encodes HSC70) and aryl hydrocarbon transcription factor activity across multiple lung ILC subsets. Overall, using an airway disease model that develops after occupational and environmental exposures, we demonstrate an effective method to better understand heterogenous ILC subset activation.

immunology↗

bioRxiv2020

miR6236, a microRNA suppressed by the anisotropic surface topography, regulates neuronal development and regeneration.

It has been well studied that the surface topography affects the growth and development of neurons. However, the precise mechanism that the surface topography leads to cellular changes remains unknown. In this study, we created an anisotropic surface using nanodiamonds and discovered this surface topography accelerates the development of primary neurons from both the central and peripheral nervous systems. Using RNA sequencing technology, a previously uncharacterized microRNA (miR6236) was found to exhibit significant and the most substantial decrease when neurons are cultured on this nanodiamond surface. Gain- and loss-of-function assays confirm that miR6236 is the predominant molecule responsible for converting the surface topography into biological responses. We further demonstrate that the depletion of miR6236 can enhance neuroregeneration on inhibitory substrate, uncovering its therapeutic potential for promoting central nervous system regeneration.

neuroscience↗

bioRxiv2019

GCNCDA: A New Method for Predicting CircRNA-Disease Associations Based on Graph Convolutional Network Algorithm

Numerous evidences indicate that Circular RNAs (circRNAs) are widely involved in the occurrence and development of diseases. Identifying the association between circRNAs and diseases plays a crucial role in exploring the pathogenesis of complex diseases and improving the diagnosis and treatment of diseases. However, due to the complex mechanisms between circRNAs and diseases, it is expensive and time-consuming to discover the new circRNA-disease associations by biological experiment. Therefore, there is increasingly urgent need for utilizing the computational methods to predict novel circRNA-disease associations. In this study, we propose a computational method called GCNCDA based on the deep learning Fast learning with Graph Convolutional Networks (FastGCN) algorithm to predict the potential disease-associated circRNAs. Specifically, the method first forms the unified descriptor by fusing disease semantic similarity information, disease and circRNA Gaussian Interaction Profile (GIP) kernel similarity information based on known circRNA-disease associations. The FastGCN algorithm is then used to objectively extract the high-level features contained in the fusion descriptor. Finally, the new circRNA-disease associations are accurately predicted by the Forest by Penalizing Attributes (Forest PA) classifier. The 5-fold cross-validation experiment of GCNCDA achieved 91.2% accuracy with 92.78% sensitivity at the AUC of 90.90% on circR2Disease benchmark dataset. In comparison with different classifier models, feature extraction models and other state-of-the-art methods, GCNCDA shows strong competitiveness. Furthermore, 10 of the top 15 circRNA-disease association candidates with the highest prediction scores were confirmed by recently published literature. These results suggest that GCNCDA can effectively predict potential circRNA-disease associations and provide highly credible candidates for biological experiments. Author SummaryThe recognition of circRNA-disease association is the key of disease diagnosis and treatment, and it is of great significance for exploring the pathogenesis of complex diseases. Computational methods can predicte the potential disease-related circRNAs quickly and accurately. Based on the hypothesis that circRNA with similar function tends to associate with similar disease, GCNCDA model is proposed to effectively predict the potential association between circRNAs and diseases by combining FastGCN algorithm. The performance of the model was verified by cross-validation experiments, different feature extraction algorithm and classifier models comparison experiments. Furthermore, 10 of the top 15 disease-associated circRNAs with the highest prediction scores were confirmed by recently published literature. It is anticipated that GCNCDA model can give priority to the most promising circRNA-disease associations on a large scale to provide reliable candidates for further biological experiment.

bioinformatics↗

bioRxiv2019

Ran GTPase regulates non-centrosomal microtubule nucleation and is transported by actin waves towards the neurite tip.

Microtubule (MT) is the most abundant cytoskeleton in neurons and controls multiple facets of their development. While the organizing center of MTs in mitotic cells is typically located at the centrosome, MT nucleation in post-mitotic neurons switches to non-centrosomal sites. A handful of proteins and organelle have been shown to promote non-centrosomal MT formation in neurons, yet the regulation mechanism remains unknown. Here we demonstrate that the small GTPase Ran is a key regulator of non-centrosomal MT nucleation in neurons. The GTP-bound Ran (RanGTP) localizes to the neurite tips and around the soma. Using the RanGTP- and RanGDP-mimic mutants, we show that RanGTP promotes MT nucleation at the tip of the neurite. To demonstrate that RanGTP can promote MT nucleation in regions other than the neurite tip, an optogenetic tool called RanTRAP was constructed to enable light-induced local production of RanGTP in the neuronal cytoplasm. An increase of non-centrosomal MT nucleation can be observed by elevating the RanGTP level along the neurite using RanTRAP, establishing a new role for Ran in regulating neuronal MTs. Additionally, the mechanism of RanGTP enrichment at the neurite tip was examined. We discovered that actin waves drive the anterograde transport of RanGTP towards the neurite tip. Pharmacological disruption of actin waves abolishes the enrichment of RanGTP and reduces the non-centrosomal MT nucleation at the neurite tip. These observations provide a novel regulation mechanism of MTs and an indirect connection between the actin and MT cytoskeletons in neurons.

cell biology↗