A5025477684- Single-cell and spatial transcriptomics
- Immune cells in cancer
- RNA modifications and cancer
- Extracellular vesicles in disease
- DNA Repair Mechanisms
- Ferroptosis and cancer prognosis
- Cancer Immunotherapy and Biomarkers
- RNA and protein synthesis mechanisms
- RNA Research and Splicing
- Cancer Genomics and Diagnostics
- Effects and risks of endocrine disrupting chemicals
- Cancer-related molecular mechanisms research
- Tryptophan and brain disorders
- Health, Environment, Cognitive Aging
- Glioma Diagnosis and Treatment
- Selenium in Biological Systems
- CRISPR and Genetic Engineering
- Reproductive System and Pregnancy
- PARP inhibition in cancer therapy
- Renal and related cancers
- Silk-based biomaterials and applications
- Epigenetics and DNA Methylation
- Trace Elements in Health
- Machine Learning in Bioinformatics
- Advanced Image and Video Retrieval Techniques
Tianjin Medical University Cancer Institute and Hospital
2020-2025
Zhejiang University
2010-2024
National Clinical Research
2024
Yale University
2024
Women's Hospital, School of Medicine, Zhejiang University
2024
Harvard University
2024
Second Affiliated Hospital of Zhejiang University
2021-2023
Zhejiang University-University of Edinburgh Institute
2021-2023
Ningxia Medical University
2020-2022
Abstract Critical-sized bone defects often lead to non-union and full-thickness of the calvarium specifically still present reconstructive challenges. In this study, we show that neurotrophic supplements induce robust in vitro expansion mesenchymal stromal cells, situ transplantation supplements-incorporated 3D-printed hydrogel grafts promote regeneration critical-sized defects. Single-cell RNA sequencing analysis reveals a unique atlas stem/progenitor cells is generated during calvarial...
Abstract Primary ovarian insufficiency (POI) is a clinical syndrome of dysfunction characterized by premature exhaustion primordial follicles. POI causes infertility, severe daily life disturbances and long-term health risks. However, the underlying mechanism remains largely unknown. We previously identified Basonuclin 1 ( BNC1 ) mutation from large Chinese pedigree found that mice with targeted Bnc1 exhibit symptoms POI. In this study, we plays key roles in reserve maintaining lipid...
Exponential accumulation of single-cell transcriptomes poses great challenge for efficient assimilation. Here, we present an approach entitled generative pretraining from (tGPT) learning feature representation transcriptomes. tGPT is conceptually simple in that it autoregressive models the ranking a gene context its preceding neighbors. We developed with 22.3 million and used four datasets to evalutate performance on analysis tasks. In addition, examine applications bulk tissues. The...
Ischemic stroke is a leading cause of mortality and disability. Diabetes mellitus, characterized by hyperglycemia, common concomitant disease ischemic stroke, which associated with autophagy dysfunction blood‑brain barrier (BBB) damage following cerebral ischemia/reperfusion (I/R) injury. At present, there no effective treatment strategy for the disease. The purpose present study was to explore molecular mechanisms underlying protective effects selenium on BBB I/R injury in hyperglycemic...
Protein-RNA interactions play pivotal roles in regulating transcription, translation, and RNA metabolism. Characterizing these offers key insights into dysregulation mechanisms. Here, we introduce Reformer, a deep learning model that predicts protein-RNA binding affinity from sequence data. Trained on 225 enhanced cross-linking immunoprecipitation sequencing (eCLIP-seq) datasets encompassing 155 RNA-binding proteins across three cell lines, Reformer achieves high accuracy predicting at...
Abstract Cartilage damage affects millions of people worldwide. Tissue engineering strategies hold the promise to provide off‐the‐shelf cartilage analogs for tissue transplantation in repair. However, current hardly generate sufficient grafts, as tissues cannot maintain size growth and cartilaginous phenotypes simultaneously. Herein, a step‐wise strategy is developed fabricating expandable human macromass (macro‐cartilage) 3D condition by employing polydactyly chondrocytes screen‐defined...
Our understanding of full-thickness endometrial regeneration after injury is limited by an incomplete molecular characterization the cell populations responsible for organ functions. To help fill this knowledge gap, we characterized 10,551 cells normal human uterine from two menstrual phases (proliferative and secretory phase) using unbiased single RNA-sequencing. We dissected heterogeneity main types (epithelial, stromal, endothelial, immune cells) full thickness tissues, population...
Abstract Protein-RNA interactions play an essential role in the regulation of transcription, translation, and metabolism cellular RNA. Here, we develop Reformer, a deep learning model that predicts protein-RNA binding affinity purely from sequence. We developed Reformer with 155 RNA protein (RBP) targets 3 cell lines. achieved high prediction accuracy at single-base resolution when tasking inferring protein- cell-type-specific affinity. conducted electrophoretic mobility shift assays to...
The rapidly evolving realm of single-cell transcriptomics offers vital new perspectives into the understanding intra- and inter-cellular molecular dynamics governing development, physiology, pathogenesis. Deep learning, a recent artificial intelligence advance with promising application for
Stem cell-based tissue regeneration strategies are promising treatments for severe endometrial injuries. However, there few appropriate seed cells regenerating a full-thickness endometrium, which mainly consists of epithelia and stroma. Müllerian ducts in female embryonic development develop into Hence, we first generated human pluripotent stem (hPSC)-derived duct-like (MDLCs) using defined effective protocol. The MDLCs bi-potent, can gradually differentiate epithelial stromal cells,...
Abstract Foundation models have demonstrated exceptional efficacy across diverse downstream tasks. However, within the realms of genomics and transcriptomics, a notable gap persists in availability that afford comprehensive understanding nucleotide sequence principles various species. Here, we present OmniNA, foundation generative model designed for learning. The was pre-trained on 91.7 million sequences corresponding annotations encompassing 1076.2 billion bases 197 words spanning multitude...
Functional coactivation between human brain regions is partly explained by white matter connections; however, how the structure-function relationship varies function remains unclear. Here, we reference large data repositories to compute maps of correspondence across hundreds specific functions and regions. We use natural language processing accurately predict for identify macroscale gradients that correlate with as well cortical thickness. Our findings suggest unfolds along a sensory-fugal...
We developed Miscell, a self-supervised learning approach with deep neural network as latent feature encoder for mining information from single-cell transcriptomes. demonstrated the capability of Miscell canonical analysis tasks including delineation clusters and identification cluster-specific marker genes. evaluated along three state-of-the-art methods on heterogeneous datasets. achieved at least comparable or better performance than other by significant margin variety clustering metrics...
Computational pathology for gigapixel whole-slide images (WSIs) at slide level is helpful in disease diagnosis and remains challenging. We propose a context-aware approach termed WSI inspection via transformer (WIT) slide-level classification holistically modeling dependencies among patches on WSI. WIT automatically learns feature representation of by aggregating features all image patches. evaluate performance state-of-the-art baseline method. achieved an accuracy 82.1% (95% CI,...
Gastric cancer is the fifth most common type of human and third leading cause cancer-related death. The purpose this study to investigate immune infiltration signatures gastric their relation prognosis. We identified two distinct subtypes (C1/C2) characterized by different signatures. C1 featured resting, epithelial-mesenchymal transition, angiogenesis pathways, while C2 enrichment MYC target, oxidative phosphorylation, E2F target pathways. subtype has a better prognosis than (HR = 0.61, 95%...
Background: Spinal cord injury (SCI) is a highly lethal and debilitating disease with variety of etiologies. To date, there no effective therapeutic modality for complete cure. The pathological mechanisms spinal at the molecular gene protein expression levels remain unclear. Methods: This study used single-cell transcriptomic analysis microarray to analyzes changes in profiles cells secretion inflammatory factors respectively, around lesion site rat SCI model. Results: Single-cell found that...
Abstract Exponential accumulation of single-cell transcriptomes poses great challenge for efficient assimilation. Here, we present an approach entitled tGPT towards integration 22.3 million by modeling gene expression rankings as generative pretraining task. is conceptually simple in that it autoregressively models the ranking a context its preceding neighbors. We demonstrated high performance on range fundamental analysis tasks and novel applications bulk tissues. The clusters cell lineage...
Background Brain tumor ranks as the most devastating cancer type. The complex immune microenvironment prevents brain from receiving therapeutic benefits. purpose of this study was to stratify tumors based on their distinct infiltration signatures facilitate better clinical decision making and prognosis prediction. Methods We developed a deep learning model characterize transcriptome. applied distill expression transcriptome samples. performed molecular subtyping with extracted unveil...