Yungang Xu

Understanding how local signalling shapes cell-state transitions requires linking spatial context to transcriptional dynamics. Here we present CellVocal, a framework that integrates spatial RNA velocity with cell-cell communication to decompose cell-state transitions into intrinsic and communication-associated components and further attribute their contributions to signalling modules and neighbouring sender cells. CellVocal recovered complex trajectories and communication perturbations in simulations and reconstructed coherent dynamics and spatially restricted ligand-receptor programmes across the mouse brain, gut and lung. Further analyses linked inferred NCAM1-NCAM1 signalling to lower cell-cycle-associated RNA velocity during embryonic neurogenesis. In melanoma, inferred local MIF-CD74+CD44 signalling corresponded to a predicted decline in the melanocytic programme. Analysis of acute liver failure identified hepatocyte-derived ANGPTL4-SDC1 signalling in association with acetaminophen-related communication enrichment and predicted matrix-remodelling RNA dynamics. By converting static ligand-receptor maps into velocity-constrained networks of signalling influence, CellVocal provides a general strategy for generating spatially resolved hypotheses about how intercellular communication contributes to predicted transcriptional changes.

Long non-coding RNAs (lncRNAs) play diverse regulatory roles in cancer, yet their functional interpretation remains largely based on molecular associations and computational predictions, while experimentally derived evidence is fragmented across heterogeneous resources. Here, we present lncRNAfunc 2.0, an updated knowledgebase for systematic functional interpretation of cancer lncRNAs through multi-source evidence integration. Compared with the previous version, lncRNAfunc 2.0 reprocesses pan-cancer data from The Cancer Genome Atlas, incorporates normal-tissue and cancer-cell-line contexts from GTEx and CCLE, substantially expands binding and interaction evidence, and introduces perturbation evidence for lncRNAs and their regulators. These resources are integrated with association-based analyses and computational predictions as complementary evidence for interpreting lncRNA-associated regulatory relationships. lncRNAfunc 2.0 organizes functional knowledge into six biological dimensions: dysregulation, upstream regulation, lncRNA-mediated mRNA regulation, transcript remodeling, protein regulation, and tumor phenotypes. This framework connects regulatory determinants of lncRNAs with their downstream molecular relationships and phenotypic associations, including mRNA regulation, alternative splicing, protein regulation, survival, molecular and immune phenotypes, and drug response. Overall, lncRNAfunc 2.0 provides an evidence-integrated framework for exploring context-dependent lncRNA functions and regulatory mechanisms across cancers. lncRNAfunc 2.0 is freely available at https://lncrnafunc.xu-bioinfo.com/.

Metabolic remodeling is a fundamental mechanism that underlies cardiomyopathy and heart failure. Preclinical and clinical studies in diverse types of heart disease have provided evidence for the activation of the myocardial Hippo pathway and that cardiomyocyte-specific Hippo pathway activation or YAP-TEAD1 inactivation drives cardiomyocyte mitochondrial damage, metabolic abnormalities and fibrosis. Here we studied whether intervention of myocardial lactate metabolism have therapeutic effect on cardiac fibrosis in cardiomyopathy models. In a Hippo pathway-activated mouse model (Mst1-cTG), single-nucleus RNA-sequencing (snRNA-seq) revealed activated fibroblasts while cardiomyocytes displayed downregulation of gene sets of mitochondrial metabolism and augmented HIF-1 signaling. Seahorse assay in isolated cardiomyocytes from adult TG mice showed impaired oxidative respiratory capacity together with enhanced glycolysis and lactate production. In TG mouse myocardium, fibroblast histone H3 lysine 18 lactylation (H3K18la), regulated by p300, was elevated starting from young age. By CUT&Tag, H3K18la was found to be enriched in promoter regions of numerous fibrotic genes with enhanced transcription. Treatment of TG mice with inhibitors to suppress either glycolysis (PX-478 as HIF-1α inhibitor) or lactate production (FX-11 as LDHA inhibitor) reduced both myocardial lactate content and H3K18la level, and ameliorated cardiac fibrosis and dysfunction. Similar efficacy was achieved in vitro or in vivo by fibroblast p300 knockdown using siRNA or AAV9. Treatment with FX-11 was similarly effective in inhibiting histone lactylation and fibrosis in mice subjected to chronic ischemia-reperfusion. Finally, we integrated our snRNA-seq and CUT&Tag studies to identify RUNX2 as a crucial transcription factor that promotes fibrotic gene expression in concert with H3K18la levels. Treatment with RUNX2 inhibitor CADD522 in TG mice improved heart dysfunction and fibrosis. In the setting of myocardial Hippo pathway activation, inhibiting the lactate-histone lactylation-RUNX2 axis effectively attenuated cardiac fibrosis.

Inferring early cell fate from single-cell RNA-sequencing data is essential for identifying cellular origins and fate plasticity in development and disease. However, existing methods often fail to exploit tree-structured lineage trajectories, limiting the accuracy and interpretability of fate mapping. Here we present DyMoTree, a computational framework that models cell fate decisions as nonlinear mappings between progenitor and terminal cell states under explicit lineage constraints. By integrating lineage graphs with a tree-structured neural architecture, DyMoTree learns lineage-resolved cell-state transition maps from single-cell transcriptomes, enabling robust inference of early fate bias and identification of fate-specific progenitor substates and driver genes. Across simulations, lineage-tracing experiments, and in vivo systems, DyMoTree outperformed existing methods in resolving early fate biases. Applications to mouse embryogenesis, lung adenocarcinoma progression, and CAR-T immunotherapy revealed regulatory programs underlying developmental and disease-associated transitions. DyMoTree provides a general framework for modeling lineage-resolved cell-state dynamics underlying development and disease progression. Abstract Figure:

Cell–cell communication coordinates tissue development, homeostasis, and immunity, yet defining signaling interactions within intact tissues remains challenging. Single-cell transcriptomics enables systematic ligand– receptor inference, but tissue dissociation removes spatial context and obscures local and region-specific signaling. Spatial transcriptomics and spatial multi-omics can recover communication in situ, although existing methods often incompletely integrate heterogeneous data or produce poorly interpretable ligand–receptor lists. Here we present CellSTIC, a framework that resolves cell–cell communication in spatial multi-omics as structured programs grounded in tissue architecture. CellSTIC integrates multimodal evidence from local neighborhoods and organizes interactions into a hierarchical semantic representation that remains traceable to underlying molecules. This enables analysis from individual ligand–receptor pairs to functional modules comparable across tissues, regions, and states. In simulations and diverse tissue datasets, CellSTIC recovered spatially coherent communication structures and domains, revealing immune, brain, developmental, and regenerative programs, and providing a general approach for generating mechanistic hypotheses in situ. Abstract Figure:

Functional perturbations of genes do not always cause expression changes, but can manifest through network rewiring or context-specific shifts in regulatory activity. However, inferring functional shifts in genes and linking them to the specific cell populations remains challenging, as most current scRNA-seq data analysis focuses either on differential gene expression or on cell abundance/state changes, but rarely associate gene perturbations with particular cell populations. Here we present scDNS, a framework that quantifies gene-specific functional perturbations by measuring information-theoretic divergence between condition-specific gene interaction network configurations. In simulated stress tests and multiple experimental datasets, scDNS prioritizes key regulators and perturbed cell populations, even when expression changes are minimal but network rewiring is pronounced. Applications to immunodeficiency mutations, stimulus responses, and viral infection reveal hidden regulatory programs and heterogeneous responder cell states. In pancreatic cancer, scDNS nominates TIMM44 as a mitochondrial sensitizer enhancing gemcitabine efficacy. Together, scDNS provides a powerful tool for inferring dynamic gene perturbations in single cells.

Advances in single-cell and spatial technologies have transformed the dissection of cell composition and tissue architecture in complex biological systems. However, identifying rare cells critical to disease pathogenesis and biological processes remains difficult, as their low abundance often masks them among dominant cell populations. We introduce RareQ, a fast and scalable framework for rare cell detection by evaluating cliquishness of each cell’s k -nearest neighborhood using single-cell omics data. Extensive benchmarking on diverse simulated and real datasets demonstrates that RareQ exceeds existing methods in accuracy, sensitivity, and efficiency. RareQ also excels at identifying both modality-specific and shared rare cells. Its versatility across various biological contexts enables the discovery of functionally distinct rare cells with unique molecular signatures in physiological and pathological context. RareQ’s application to spatial transcriptomics data reveals anatomically distinct and clinically relevant rare cell populations. Together, RareQ offers an efficient approach for rare cell discovery, enhancing insights into tissue organization and disease mechanisms.

Extrachromosomal circular DNAs (eccDNAs) are closed circular DNA molecules widespread across eukaryotic cells, with emerging roles in gene regulation and tumor progression. Experimental assays remain costly and incomplete, underscoring the need for computational approaches. To address this, a deep learning framework termed DeepECC has been established to overcome the challenges posed by eccDNA heterogeneity and its complex biogenesis. Through a two-stage training strategy, DeepECC models the local sequence context flanking both the start and end breakpoints, thereby capturing mechanistically informative features that are often overlooked when analyses focus solely on eccDNA body sequences. Applied to multi-species (human, mouse, gallus) datasets, DeepECC robustly captures conserved breakpoint features, with a marked preference for GC-rich and transcriptionally active regions. Genome-wide scanning reveals non-uniform distributions of human cancer eccDNAs enriched in enhancers, expression quantitative trait loci, and CTCF sites, suggesting regulatory functions in tumor progression. Motif analysis further implicates ribosomal activity, translational regulation, and DNA damage response. Furthermore, genome-wide eccDNA predictions are integrated into the UCSC Genome Browser, enabling convenient querying and visualization of cancer-related eccDNAs associated with specific genes or genomic regions, facilitating functional interpretation for experimental research. Collectively, DeepECC provides a generalizable framework for systematic eccDNA discovery and insights into their functional significance in cancer.

RNA modifications represent a dynamic layer of gene expression regulation, RNA stability, and translation with profound implications for cellular function and disease. However, the critical regulation and functions of RNA-modifying proteins (RMPs) remain poorly understood. Here, we present a large-scale characterization of RMPs through 378 multiomics datasets encompassing genomics, bulk and single-cell transcriptomics, epitranscriptomics, proteomics, and posttranslational modifications (PTMs) across 63 human tissues. Our analysis of experimental perturbations of RMPs revealed dynamic differential modification peaks and expressed genes. We applied nonnegative matrix factorization to annotate RMP-mediated cell types in single-cell transcriptomes. Functional annotations in acute myeloid leukemia (AML) revealed RMPs such as ALKBH5 as critical mediators of m6A dynamics, influencing pathways involved in translation initiation, immune regulation, and tumorigenesis. We revealed cell type-specific modification patterns, including those in ALKBH5-enriched AML stem cells with special ligand‒receptor interactions and genetic variations modulated by m6A. We integrated proteogenomic data to uncover PTM-associated regulatory, mutation, and protein‒protein interaction networks linked to RMPs. We developed RMzyme, a platform that consolidates our findings and provides insights into RMPs and their downstream effects. This resource is expected to facilitate biomedical research into the molecular mechanisms of human diseases through the lens of RNA modifications and multiomics data integration.