Latest in Science

Abstract: RNA velocity techniques have emerged as efficient tools for unraveling the complex trajectories of cell development and differentiation. However, most of existing RNA velocity approaches are constrained by estimating transcriptional parameters for each gene in isolation and neglects the regulatory relationships among genes, which limits the ability to jointly resolve the dynamic rewiring of gene regulation and the underlying gene transcriptional kinetics across cell state transitions. To address these limitations, we present GRAVITY, a novel deep learning framework that explicitly integrates regulatory dynamics into transcriptional kinetics inference and utilizes a refined two-stage optimization strategy. Benchmarking across various simulated and real single-cell RNA sequencing datasets demonstrates that GRAVITY accurately infers both cellular and gene trajectories, along with their associated kinetic parameters. Most importantly, GRAVITY uncovers terminal cell states L5/L6 in embryonic brain development dataset. Furthermore, GRAVITY not only provides mechanistic insights by identifying the driver regulatory factors and modules governing cell fate, but also enables the systematic in silico simulation of cellular velocity changes induced by targeted regulatory perturbations.

Abstract: Despite the fact that microbes in natural environments spend most of their time in growth arrest, we understand little about how this physiological state affects the performance of engineered genetic circuits. Here, we measure repression curves from a library of genetic NOT gates at single-cell resolution in Escherichia coli under both active growth and growth arrest to systematically investigate how growth arrest affects circuit behavior. We find that the impact of growth arrest on circuit performance is almost entirely dominated by a single effect: a >100-fold reduction in unrepressed expression levels. Growth arrest caused gene expression noise to increase moderately and had only minimal impacts on the sensitivity and sharpness of the repression curves. Our work shows both that conventional genetic circuit design paradigms are currently insufficient to develop circuits that can function properly under growth arrest, but also that addressing the reduction in just a single performance parameter would be sufficient to resolve this problem. This work expands our understanding of bacterial gene regulation under growth arrest and lays the groundwork for new design paradigms that will be essential in ensuring the safe and reliable performance of synthetic biology systems in real-world environments.

Abstract: DNA methylation is a central epigenetic modification that regulates gene expression, maintains genomic stability, and guides cellular differentiation. However, direct measurements of DNA methylation, such as whole genome bisulfite sequencing or DNA methylation arrays, are costly and require substantial DNA input, limiting their scalability for large cohorts and their applicability to emerging modalities such as single cell and spatially resolved transcriptomics. In this study, motivated by the fact that DNA methylation is fundamentally a metabolic process, we investigate whether sample-wise DNA methylation activity can be inferred directly from transcriptomic profiles of genes involved in methionine and one-carbon metabolism. We show that a compact metabolic model comprising seven core reaction steps and 98 genes accurately predicts total DNA methylation activity across matched transcriptomic and methylation datasets from CCLE, TCGA, GTEx, and an independent single-cell multi-omics data set. Building on this framework, we develop Total DNA Methylation Activity (TDMA), a physics-informed neural network based score that enables robust estimation of DNA methylation activity from bulk, single-cell, and spatial transcriptomics data. We demonstrate that TDMA captures methylation-dependent transcriptional regulation and identifies genes and pathways under epigenetic control. Applying TDMA to GTEx, we further reveal strong associations between the predicted total methylation activity, chronological aging, and established epigenetic clocks. We also demonstrated that TDMA can serve as a transcriptomics-derived epigenetic clock and highlights age-dependent roles of folate and glutathione metabolism in epigenetic aging. Applying TDMA to single cell and spatial transcriptomics data collected from pancreatic adenocarcinoma (PDAC), we identified that methionine metabolism and DNA methylation regulates T cell cytotoxicity in the tumor microenvironment of PDAC. Together, this work establishes a scalable, modality-agnostic framework for estimating DNA methylation activity from transcriptomics and provides new insights into the metabolic regulation of epigenetic aging.

Abstract: Technologies developed over the past decade have made Saccharomyces cerevisiae a promising platform for producing various natural products. Balancing multi-enzyme expression, while maintaining robust microbial growth, remains a limiting factor for engineering long biosynthetic pathways in yeast. Here, we improved the transcriptional capacity of our previously developed isopropyl {beta}-D-1-thiogalactopyranoside (IPTG)-inducible synthetic transcription factors (synTFs) derived from the plant JUB1 DNA-binding domain. To this end, at cysteine positions within surface-exposed loop regions of a JUB1-derived DNA-binding scaffold, we introduced a short peptide to enhance loop flexibility while providing local stability and orientation. The generated synTFs, so-called JUB1-X synTFs, varying in strength, have been successfully used to improve the synthesis of 3'-phosphoadenosine 5'-phosphosulfate (PAPS), a universal sulfate donor necessary for the synthesis of bioactive molecules, including therapeutic glycosaminoglycans and sulfolipids. Using only this engineered yeast strain in simple batch culture, PAPS accumulation of 21.4 {+/-} 5.8 mg/g cdw was achieved after only 5 hours of inducing the expression of JUB1-X synTFs. Beyond PAPS production, the design principle demonstrated here provides a generalizable strategy to fine-tune other plant-derived synTFs, expanding the regulatory capabilities of existing synTF collections. Together, this work offers a modular, scalable approach to constructing high-performance gene circuits and supports the development of yeast cell factories for complex metabolic and synthetic biology applications.

Abstract: High-resolution histology images are indispensable for pathology and increasingly serve as the structural backbone for spatial omics. Yet whole-slide images (WSIs) frequently contain artifacts, acellular voids, and background regions that, when included in computational workflows, introduce noise, degrade model accuracy, and compromise biological interpretation. Existing tools provide only coarse foreground-background separation, leaving a gap in fine-grained quality control (QC). Here we present HistoSweep, a scalable framework that generates morphology-aware tissue masks at cellular resolution. By integrating density filtering, texture descriptors, and adaptive thresholding, HistoSweep systematically removes non-informative tissue regions while preserving biologically meaningful microstructures. It processes billion-pixel WSIs in minutes on standard CPUs, requiring no GPU acceleration, and is deployable across research and clinical settings. Across 25 WSIs spanning distinct tissues, disease states, and spatial omics platforms, HistoSweep consistently outperformed existing methods. It enhanced visualization and segmentation, improved virtual cell type predictions, and safeguarded spatial transcriptomics integrity by detecting transcript leakage and transcript-histology misalignment. By enabling fine-grained, scalable QC, HistoSweep provides a foundational preprocessing step for reliable and reproducible digital pathology and spatial omics analyses.

Abstract: Motivation: Current machine learning methods for enzyme function prediction primarily treat proteins as independent entities, ignoring the metabolic context in which they operate. This reductionist approach often generates biologically implausible annotations that fail to satisfy stoichiometric or thermodynamic constraints. While genome-scale metabolic models (GEMs) enforce systemic coherence, traditional reconstruction workflows decouple functional annotation from network assembly, resulting in information loss during the discretization of enzymatic evidence. Results: We developed METEOR, a framework that integrates continuous confidence scores from deep learning models directly into a Mixed-Integer Linear Programming (MILP) formulation. By conditioning predictions on global metabolic requirements, METEOR resolves functional ambiguities where sequence signals are weak. Evaluation on the Price-149 dataset shows that our method yields physiologically viable networks capable of sustaining growth and improve protein-level predictions over baseline methods. Validation on 6,894 bacterial genomes reveals that system-aware refinement increases the recall of experimentally observed phenotypes substantially. Furthermore, we show that constraint-driven selection can still improve annotation performance in highly incomplete genomes. Our results suggest that functional annotation should be treated as a unified inference problem where global system constraints supervise local predictions.

Abstract: The effects of perturbation on a biological system can be readily measured in terms of transcriptional changes. However, despite a wealth of transcriptional perturbation response data, there are currently few methods to draw equivalence between the many biological systems used to generate that data and a specific system of interest. Here we use density analysis of transcriptional correlations to computationally predict whether a given perturbation readout is relevant to Stem Cell derived islet (SC-Islet) maturation. The approach, Density Based Self-Organizing Map Analysis (DBSOMA), first learns patterns of gene expression represented in scRNA-seq sets by clustering genes with the Self-Organizing-Map (SOM) algorithm. Perturbation expression profiles and other gene lists are then projected onto the SOM grid, where the degree of clustering is determined by the Density-Based Spatial Clustering of Applications with Noise (DBSCAN) algorithm. We applied DBSOMA to SC-Islet maturation and identified known and novel regulators of {beta}-cell maturation. This workflow can be applied broadly to biological systems where single-cell RNA-sequencing data is available, and a desired outcome can be represented in transcriptional changes.

Abstract: Conditional and joint (COJO) analysis of genome-wide association study (GWAS) summary statistics to identify single nucleotide polymorphisms (SNPs) independently associated with a trait is standard in post-GWAS pipelines. GWAS meta-analyses are increasingly conducted across multiple ancestry groups but how to perform COJO in a multi-ancestry context is not known. Here we introduce Manc-COJO, a method for multi-ancestry COJO analysis. Simulations and real-data analyses show that Manc-COJO improves the detection of independent association signals and reduces false positives compared to COJO and ad hoc adaptations for multi-ancestry use. We also introduce Manc-COJO:MDISA, a follow-up within ancestry algorithm to identify ancestry-specific associations after fitting Manc-COJO identified SNPs. The C++ implementation of Manc-COJO substantially improves on computational efficiency (for single ancestry >120 times faster than GCTA-COJO software) and supports linkage disequilibrium references derived either from individual-level genotype data or pre-computed matrices, facilitating analysis when data sharing is limited.

Abstract: Although cell type annotation has become an integral part of single-cell analysis workflows, the assessment of computational annotations remains challenging. Many annotation tools transfer labels from an annotated reference dataset to a new query dataset of interest, but blindly transferring labels from one dataset to another has its own set of challenges. Often enough there is no perfect alignment between datasets, especially when transferring annotations from a healthy reference atlas for the discovery of disease states. We present scDiagnostics, a new open-source software package that facilitates the detection of complex or ambiguous annotation cases that may otherwise go unnoticed, thus addressing a critical unmet need in current single-cell analysis workflows. scDiagnostics is equipped with novel diagnostic methods that are compatible with all major cell type annotation tools. We demonstrate that scDiagnostics reliably detects complex or conflicting annotations using both carefully designed simulated datasets and diverse real-world single-cell datasets. Our evaluation demonstrates that scDiagnostics reliably identifies misleading annotations that systematically distort downstream analysis and interpretation and that would otherwise remain undetected. The scDiagnostics R package is available from Bioconductor (https://bioconductor.org/packages/scDiagnostics).

Abstract: Combined small-cell lung cancer (cSCLC) is a rare and aggressive subtype of small-cell lung cancer (SCLC) characterized by mixed histology comprising SCLC and non-small cell lung cancer (NSCLC) or large cell neuroendocrine carcinoma (LCNEC) components. Despite its histological heterogeneity and even poorer prognosis than de novo SCLC, cSCLC is clinically managed as pure SCLC, largely due to the lack of molecular insights into its biology, lineage plasticity, and tumor microenvironment (TME). Here, we perform multi-omics profiling, including spatially-resolved whole-exome sequencing (WES), spatial transcriptomics (ST) and single-nucleus RNA sequencing (snRNA-seq), across 19 treatment-naive cSCLC tumors spanning all major histological subtypes. Our analysis reveals that SCLC and NSCLC/LCNEC components share a monoclonal origin, with histological divergence characterized by distinct mutation and copy number alteration patterns. ST and snRNA-seq uncover spatially exclusive or interspersed tumor domains, with distinct TME compositions and immune landscapes. Notably, fibroblast-rich regions enriched for an aggressive fibroblast subtype form boundaries between tumor domains, potentially influencing immune TME and treatment responses. We identify extensive lineage plasticity within cSCLC, including active LUAD-to-SCLC transdifferentiation and SCLC subtype coexistence, suggesting transitional cellular states not captured by traditional diagnostics. Leveraging these insights, we developed the cSCLC Detector, a sensitive mutation-based diagnostic assay that improves the detection of cSCLC in tissue and liquid biopsy samples. Our findings offer critical insights into cSCLC lineage plasticity, cellular evolution, and microenvironmental interactions, underscoring the need for tailored treatment strategies and diagnostic frameworks for this aggressive cancer subtype.

Abstract: The MET receptor tyrosine kinase is mutated or amplified in ~6% of non-small cell lung cancers (NSCLC) and overexpressed in NSCLC. Here, we report a novel shark-derived single-domain variable new antigen receptor (VNAR) with high MET affinity for theranostic applications. Following immunization of a juvenile nurse shark (Ginglymostoma cirratum) with the extracellular domain of human MET, we identified a VNAR clone exhibiting high MET selectivity in vitro. As a bivalent human Fc fusion, vMET1-Fc was selectively internalized by MET-expressing cell lines and xenografts. Radiolabeled with zirconium-89 (Zr-89-vMET1-Fc), it enabled PET/CT detection of MET-positive NSCLC xenografts. As a therapeutic, Lu-177-vMET1-Fc delayed tumor growth in MET-mutant and MET-amplified cell line-derived xenografts. Non-human primate studies in healthy rhesus macaques confirmed favorable biodistribution, predictable clearance, and minimal off-target uptake. Together, these findings establish vMET1-Fc as a theranostic agent for imaging and treating MET-altered NSCLC.

Abstract: 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.

Abstract: Renal cell carcinoma (RCC) frequently exhibits resistance to immune checkpoint blockade, highlighting the need for strategies that enhance tumor-specific T cell priming and improve immune access to the tumor microenvironment. Here we show that vaccination targeting tumor-associated mitochondrial antigens (TAMAs), derived from tumor-specific mitochondrial DNA (mtDNA) missense mutations, synergizes with PD 1/PD L1 blockade to overcome checkpoint refractoriness in the RENCA RCC model. TAMAs vaccination elicits antigen-specific T cell responses, increases intratumoral CD8+ T cell infiltration, and reduces immunosuppressive myeloid populations, resulting in delayed tumor progression and improved survival when combined with checkpoint inhibition. In parallel, TAMAs + checkpoint blockade induces vascular remodeling characterized by increased pericyte coverage, reduced vascular leakage, improved perfusion and reduced hypoxia. Mechanistically, vascular remodeling is driven by CD8+ T cell dependent, IFN{gamma} associated immune activity and is associated with endothelial apoptosis and diminished intratumoral CD31 signal. We further identify tumor-to-endothelium mitochondrial transfer as a mechanism linking mitochondrial neoantigens to the tumor vascular compartment: tumor-derived mitochondria enter human and mouse endothelial cells in vitro and in vivo, and tumor-associated mtDNA mutations are detectable in endothelial fractions from murine tumors and human RCC specimens. Human endothelial cells can present mitochondrial neoantigens via MHC class I and become targets of TAMAs-specific CD8+ T cell cytotoxicity, including following mitochondrial acquisition from tumor cells. Together, these findings establish mitochondrial neoantigen immunity as a tractable approach to enhance checkpoint responses and reveal mitochondrial transfer as an antigenic bridge that expands immune targeting to the tumor vasculature.

Abstract: Hypoxia within the tumor microenvironment profoundly limits the efficacy of immune and cellular therapies, yet most in vitro cytotoxicity assays neglect spatial oxygen heterogeneity. We developed Coverslip Hypoxia High-Content Screening (CH-HCS), a simple, scalable 2D co-culture platform that enables quantitative, region-resolved evaluation of CAR-T cell activity across controlled oxygen gradients within a single well. In CH-HCS, a 5 mm glass coverslip placed over tumor-stroma co-cultures restricts oxygen diffusion, generating concentric hypoxia-normoxia zones in standard 96-well plates. Fluorescently labeled CD19+ Raji or CD19- K562 tumor cells, stromal cells, and anti-CD19 CAR-T cells were analyzed using multiparametric fluorescence imaging with an ImageXpress Micro XLS system coupled to custom CellProfiler-KNIME pipelines, enabling segmentation of spatial Regions of Interest (InnerCore, OuterCore, Periphery, and Outside) and single-cell quantification of tumor death (SYTOX Green) and T-cell morphodynamics. The platform reproducibly established oxygen gradients that strongly shaped cellular behavior: CAR-T cytotoxicity and motility were maximal in normoxic regions but markedly suppressed within hypoxic cores, whereas effector cell survival increased under low oxygen. Unlike bulk cytotoxicity assays, CH-HCS directly visualizes spatial functional heterogeneity within the same well, allowing simultaneous comparison of matched hypoxic and normoxic compartments. Together, CH-HCS provides a cost-effective, high-throughput, and physiologically relevant tool for preclinical screening of CAR-T products and therapeutic strategies aimed at overcoming hypoxia-driven immune resistance at the tumor-stroma interface.