Biology subjects

Yu, K.-H.

Publications and source records attributed to Yu, K.-H..

8 recordsLinked to original sources

bioRxiv2026

RECON infers regions of interest from H&E images and reconstructs whole-slide molecular profiles at single-cell resolution

Spatial omics technologies resolve molecular expression and spatial architecture at single-cell resolution, but profiling whole slides remains costly. In practice, only a few regions of interest (ROIs) are profiled, leaving the rest of the tissue unmeasured. S2-omics was the first framework to unify ROI selection with out-of-ROI prediction, but it operates on superpixels rather than individual cells and predicts discrete cell types rather than continuous molecular profiles. Superpixel-based representations do not explicitly preserve cell boundaries, while categorical cell-type labels cannot quantify molecular expression within cells. Here we present RECON, a two-stage framework that performs ROI inference and whole-slide molecular reconstruction at single-cell resolution, predicting both continuous molecular profiles and discrete cell-type labels. In the first stage, RECON extracts morphological and microenvironmental features from individual cells to identify a representative ROI for spatially resolved single-cell molecular profiling. In the second stage, RECON trains deep learning models on molecular measurements acquired within the selected ROI and reconstructs transcriptomic or proteomic profiles for all remaining cells on the slide. Benchmarked against pathologist annotations, RECONs ROI selection outperforms the superpixel-based S2-omics approaches (IoU: 0.75 versus 0.64). For transcriptomics, refining the modeling unit from superpixels to single cells improves per-gene Pearson correlation by 22%. For proteomics, RECON surpasses the current state-of-the-art method, ROSIE, across all 16 markers, with a median per-cell Pearson correlation of 0.91 versus 0.84. Moreover, RECON delineates tumour boundaries and regions with distinct immune-cell densities, and highlights candidate tertiary lymphoid structures. Together, these results demonstrate that RECON enables informative ROI selection and whole-slide molecular reconstruction at single-cell resolution for both spatial transcriptomics and spatial proteomics.

bioinformatics↗

bioRxiv2026

Electroadhesion of polymer networks by polycation interfacial bridging: sticky electrophoresis, ionic complexation, and chain entanglement

An e-GLUE is a polymer network containing interpenetrating polycations, which can bond the anionic network of mucosa through interfacial polycation bridging under an electric field. Such an electroadhesion involves electrophoresis of polycations, ionic complexation between polycations and the anionic network, and polycation-network entanglement, yet their quantitative understanding is lacking. Here, we formulate a theoretical model to describe electroadhesion of polymer networks by polycation interfacial bridging. We use a diffusion-drift model coupled with a Bell-like field-dependent chain friction to describe the sticky electrophoresis of polycations in an anionic sea. The formation of ionic bonds is determined by local availability of cations and anions over the penetration depth. To debond, a force must either pull polycations out from the e-GLUE network or first dissociate them from ionic complexes and then pull out from the anionic network. We model chain pullout from the bulk networks to the interface as a viscous drag against water. The adhesion strength is calculated by summing the debonding force for each polycation per unit area across all chains. Our model quantitatively links electric field strength, applied duration, polycation chain length, and cation concentration to polycation electrophoresis kinetics, ionic bond formation, and adhesion strength. We further conduct electroadhesion tests, and our model predicts well with the experimental data. Lastly, we discuss the use of the model to guide the e-GLUE design. TOC graphic O_FIG O_LINKSMALLFIG WIDTH=200 HEIGHT=90 SRC="FIGDIR/small/730541v1_ufig1.gif" ALT="Figure 1"> View larger version (45K): org.highwire.dtl.DTLVardef@16524c6org.highwire.dtl.DTLVardef@15163aeorg.highwire.dtl.DTLVardef@673949org.highwire.dtl.DTLVardef@e207a0_HPS_FORMAT_FIGEXP M_FIG C_FIG For Table of Contents use only

bioengineering↗

bioRxiv2026

Learning residue-level context for modeling protein-protein interactions

Protein language models (PLMs) enable prediction of protein properties by learning residue-level features from sequence, yet most PLM-based approaches to protein-protein interactions aggregate information across entire proteins, limiting resolution and interpretability. Here we present ReCLIP, a transformer-based framework that learns interaction-specific representations at the level of individual residues by combining intra-protein residue neighborhoods with residue-conditioned representations of interaction partners. We show that residue-centered context provides a general framework for modeling protein interactions across diverse biological settings. ReCLIP accurately predicts mutation-induced perturbations (AUROC = 0.973), generalizes to post-translational modifications that do not alter sequence (AUROC = 0.822), and enables zero-shot prediction of peptide-MHC binding across unseen alleles (AUROC up to 0.972). Analysis of learned residue neighborhoods reveals structurally and functionally coherent patterns aligned with known determinants of binding. Applied to clinically annotated genetic variants, ReCLIP identifies disease-associated interaction perturbations that link pathogenic variants to specific molecular interaction contexts. Our results establish a generalizable and interpretable framework for modeling protein interactions and provide insights into how residue-level context shapes interaction specificity and its perturbation.

bioinformatics↗

bioRxiv2023

SpatialCells: Automated Profiling of Tumor Microenvironments with Spatially Resolved Multiplexed Single-Cell Data

BackgroundCancer is a complex cellular ecosystem where malignant cells coexist and interact with immune, stromal, and other cells within the tumor microenvironment. Recent technological advancements in spatially resolved multiplexed imaging at single-cell resolution have led to the generation of large-scale and high-dimensional datasets from biological specimens. This underscores the necessity for automated methodologies that can effectively characterize the molecular, cellular, and spatial properties of tumor microenvironments for various malignancies. ResultsThis study introduces SpatialCells, an open-source software package designed for region-based exploratory analysis and comprehensive characterization of tumor microenvironments using multiplexed single-cell data. ConclusionsSpatialCells efficiently streamlines the automated extraction of features from multiplexed single-cell data and can process samples containing millions of cells. Thus, SpatialCells facilitates subsequent association analyses and machine learning predictions, making it an essential tool in advancing our understanding of tumor growth, invasion, and metastasis. Availability of code and materialshttps://github.com/SemenovLab/SpatialCells.

bioinformatics↗

bioRxiv2023

Bayesian modeling disentangles language versus executive control disruption in stroke

Stroke is the leading cause of long-term disability worldwide. Incurred brain damage disrupts cognition, often with persisting deficits in language and executive capacities. Despite their clinical relevance, the commonalities, and differences of language versus executive control impairments remain under-specified. We tailored a Bayesian hierarchical modeling solution in a largest-of-its-kind cohort (1080 stroke patients) to deconvolve language and executive control in the brain substrates of stroke insults. Four cognitive factors distinguished left- and right-hemispheric contributions to ischemic tissue lesion. One factor delineated language and general cognitive performance and was mainly associated with damage to left-hemispheric brain regions in the frontal and temporal cortex. A factor for executive control summarized control and visual-constructional abilities. This factor was strongly related to right-hemispheric brain damage of posterior regions in the occipital cortex. The interplay of language and executive control was reflected in two factors: executive speech functions and verbal memory. Impairments on both were mainly linked to left-hemispheric lesions. These findings shed light onto the causal implications of hemispheric specialization for cognition; and make steps towards subgroup-specific treatment protocols after stroke.

neuroscience↗

bioRxiv2022

Bayesian stroke modeling details sex biases in the white matter substrates of aphasia

Ischemic cerebrovascular events often lead to aphasia. Previous work provided hints that such strokes may affect women and men in distinct ways. Women tend to suffer strokes with more disabling language impairment, even if the lesion size is comparable to men. In 1,401 patients, we isolated data-led representations of anatomical lesion patterns and hand-tailored a Bayesian analytical solution to carefully model the degree of sex divergence in predicting language outcomes [~]3 months after stroke. We located lesion-outcome effects in the left-dominant language network that highlight the ventral pathway as a core lesion focus across different tests of language performance. We provide newly detailed evidence for sex-specific brain-behavior associations in the domain-general networks associated with cortico-subcortical pathways, with unique contributions of the fornix in women and cingular fiber bundles in men. Our collective findings suggest diverging white matter substrates in how stroke causes language deficits in women and men. Clinically acknowledging such sex disparities has the potential to improve personalized treatment for stroke patients worldwide.

neuroscience↗

bioRxiv2021

Strategic infarct locations for post-stroke depressive symptoms: a lesion- and disconnection-symptom mapping study

BackgroundDepression is the most common neuropsychiatric complication after stroke. Infarct location is associated with post-stroke depressive symptoms (PSDS), but it remains debated which brain structures are critically involved. We performed a large-scale lesion-symptom mapping study to identify infarct locations, and white matter disconnections, associated with PSDS. MethodsWe included 553 patients (age 69{+/-}11 years, 42% female) with acute ischemic stroke. PSDS were measured using the 30-item Geriatric Depression Scale (GDS-30). Multivariable support vector regression (SVR)-based analyses were performed both at the level of individual voxels (SVR-VLSM) and predefined regions of interest (SVR-ROI) to relate infarct location to PSDS. We externally validated our findings in an independent stroke cohort (N=459). Finally, disconnectome-based analyses were performed using SVR-VLSM, in which white matter fibers disconnected by the infarct were analyzed instead of the infarct itself. Results: Infarcts in the right amygdala, right hippocampus and right pallidum were consistently associated with PSDS (permutation-based p<0.05) in SVR-VLSM and SVR-ROI. External validation (N=459) confirmed the association between infarcts in the right amygdala and pallidum, but not the right hippocampus, and PSDS. Disconnectome-based analyses revealed that disconnections in the right parahippocampal white matter, right thalamus and pallidum, and right anterior thalamic radiation were significantly associated (permutation-based p<0.05) with PSDS. ConclusionsInfarcts in the right amygdala and pallidum, and disconnections of right limbic and frontal cortico-basal ganglia-thalamic circuits, are associated with PSDS. Our findings provide a comprehensive and integrative picture of strategic infarct locations for PSDS, and shed new light on pathophysiological mechanisms of depression after stroke.

neuroscience↗

bioRxiv2019

Classifying Non-Small Cell Lung Cancer Histopathology Types and Transcriptomic Subtypes using Convolutional Neural Networks

Non-small cell lung cancer is a leading cause of cancer death worldwide, and histopathological evaluation plays the primary role in its diagnosis. However, the morphological patterns associated with the molecular subtypes have not been systematically studied. To bridge this gap, we developed a quantitative histopathology analytic framework to identify the gene expression subtypes of non-small cell lung cancer objectively. We processed whole-slide histopathology images of lung adenocarcinoma (n=427) and lung squamous cell carcinoma patients (n=457) in The Cancer Genome Atlas. To establish neural networks for quantitative image analyses, we first build convolutional neural network models to identify tumor regions from adjacent dense benign tissues (areas under the receiver operating characteristic curves (AUC) > 0.935) and recapitulated expert pathologists diagnosis (AUC > 0.88), with the results validated in an independent cohort (n=125; AUC > 0.85). We further demonstrated that quantitative histopathology morphology features identified the major transcriptomic subtypes of both adenocarcinoma and squamous cell carcinoma (P < 0.01). Our study is the first to classify the transcriptomic subtypes of non-small cell lung cancer using fully-automated machine learning methods. Our approach does not rely on prior pathology knowledge and can discover novel clinically-relevant histopathology patterns objectively. The developed procedure is generalizable to other tumor types or diseases.

pathology↗