Biology subjects

Aronow, B. J.

Publications and source records attributed to Aronow, B. J..

16 recordsLinked to original sources

bioRxiv2026

A Conversational Multi-Agent AI System for Integrated Multi-Omics Analysis and Biomedical Discovery

Single-cell and spatial omics offer unprecedented opportunities to decipher the mechanisms of disease, however, this process requires teams of experts, iterative trial-and-error and reasoning across modalities. Here we present LungChat (https://chat.lungmap.net), a conversational system for integrated multi-omics analysis and biomedical discovery, deployed as a hierarchical multi-agent architecture in which a supervisor decomposes natural-language questions into parallel, tool-grounded tasks spanning single-cell and spatial analyses, literature and clinical-trial synthesis, and drug repurposing. To predict new therapeutics, LungChat implements Direction-Aware Repurposing and Targeting (DART) to distinguish perturbations that reverse disease transcriptional programs from those that reinforce them, at the cell-type level, for safety prediction. Controlled architecture ablations showed that hierarchical orchestration improved grounded abstention and token efficiency and preserved strong performance on complex multi-step tasks. In pulmonary disease case studies, LungChat independently prioritized saracatinib for IPF through drug-connectivity screening, followed by DART-based cell-type analysis; the same compound has been evaluated in the STOP-IPF clinical trial (NCT04598919). The system also recovered fluticasone propionate, an established COPD therapy, through a single orchestrated analysis. This tissue-agnostic system provides a blueprint for verifiable agentic AI systems that support reproducible scientific discovery.

bioinformatics↗

bioRxiv2026

HLiCA: An integrated cell atlas of the healthy human liver

The human liver is composed of a heterogeneous mix of cell types. How these distinct populations contribute individually and collectively to liver function remains poorly understood. Although single-cell technologies have advanced our understanding of liver biology, individual studies have often been limited by small donor cohorts and inconsistent cell type annotations. Integrating multiple datasets can overcome these challenges and better capture biological variability. We present the Human Liver Cell Atlas (HLiCA), an integrated reference of non-disease liver cells assembled from eight datasets across six research centers, encompassing more than 525,000 cells from 110 donors. Developed in collaboration with the Human Cell Atlas Liver Bionetwork, the HLiCA incorporates expert-curated cell annotations refined through community feedback and dedicated cell type annotation meetings. The HLiCA classifies cells into six lineages and expands the cell type resolution to include 47 distinct cell types. Starting from raw sequencing reads, we realigned all data and performed rigorous benchmarking to ensure robust integration across technical and biological variables. Genetic ancestry was inferred for all samples to evaluate the range of ancestral backgrounds represented in the atlas. The expanded cell type annotation enabled identification of previously unrecognized liver cell types, including NRXN1+ stromal cells. Their presence was validated using spatial transcriptomics, which localized NRXN1+ stromal cells to periportal regions. With the number of donors included in the HLiCA we were able to examine cell type specific associations with demographic covariates. In hepatocytes, drug metabolism genes showed differential expression between sexes, and in cholangiocytes, mucus-production genes varied with age. As the largest and most genetically diverse human liver cell atlas to date, the HLiCA provides a comprehensive, well-annotated reference for the field, annotated by expert consensus. This resource will enable deeper interrogation of liver cellular diversity, architecture, and function in the healthy human liver and serve as a reference to understand changes that occur with disease.

genomics↗

bioRxiv2025

Oncogenic Role of Aberrant EZH2 in Hepatoblastoma

Hepatoblastoma (HB) is the most common pediatric liver malignancy. However, its cellular origin and molecular drivers remain poorly defined. Using single-nuclear RNA sequencing (snRNA-seq), we identified a proliferative, hepatocyte-derived tumor cell population (cycling HepT) enriched for Enhancer of Zeste Homolog 2 (EZH2) expression, particularly in the aggressive embryonal subtype. Integrative genomic and transcriptomic profiling confirmed EZH2 overexpression. Disruption of the PRC2 complex was evident through mislocalization and reduced expression of SUZ12, a core component. EZH2 overexpression correlated with upregulation of mitotic regulators such as AURKB and Ki67 in human HB gene expression analysis as compared to background liver. Targeted sequencing identified variants of uncertain significance in EZH2 and SUZ12 in 11 of 11 patient tumors. Pharmacologic inhibition of EZH2 with EPZ-6438 reduced proliferation and sensitized HB cells to cisplatin through gene regulation, potentially modulating platinum accumulation both in vitro and in vivo. In summary, EZH2 promotes HB progression through epigenetic silencing and noncanonical signaling pathways. These findings support EZH2s contribution to HB pathogenesis, therefore identifying it as a novel therapeutic target. STATEMENT OF SIGNIFICANCEEZH2 is enriched in hepatoblastoma, suggesting its role in early tumorigenesis. Its inhibition, combined with cisplatin, produces synergistic anti-tumor effects, supporting EZH2 as both an oncogenic and therapeutic target.

cancer biology↗

bioRxiv2023

Bering: joint cell segmentation and annotation for spatial transcriptomics with transferred graph embeddings

Single-cell spatial transcriptomics such as in-situ hybridization or sequencing technologies can provide subcellular resolution that enables the identification of individual cell identities, locations, and a deep understanding of subcellular mechanisms. However, accurate segmentation and annotation that allows individual cell boundaries to be determined remains a major challenge that limits all the above and downstream insights. Current machine learning methods heavily rely on nuclei or cell body staining, resulting in the significant loss of both transcriptome depth and the limited ability to learn latent representations of spatial colocalization relationships. Here, we propose Bering, a graph deep learning model that leverages transcript colocalization relationships for joint noise-aware cell segmentation and molecular annotation in 2D and 3D spatial transcriptomics data. Graph embeddings for the cell annotation are transferred as a component of multi-modal input for cell segmentation, which is employed to enrich gene relationships throughout the process. To evaluate performance, we benchmarked Bering with state-of-the-art methods and observed significant improvement in cell segmentation accuracies and numbers of detected transcripts across various spatial technologies and tissues. To streamline segmentation processes, we constructed expansive pre-trained models, which yield high segmentation accuracy in new data through transfer learning and self-distillation, demonstrating the generalizability of Bering.

bioinformatics↗

bioRxiv2023

Spatial transcriptome profiling uncovers metabolic regulation of left-right patterning

Left-right patterning disturbance can cause severe birth defects, but it remains least understood of the three body axes. We uncovered an unexpected role for metabolic regulation in left-right patterning. Analysis of the first spatial transcriptome profile of left-right patterning revealed global activation of glycolysis, accompanied by right-sided expression of Bmp7 and genes regulating insulin growth factor signaling. Cardiomyocyte differentiation was left-biased, which may underlie the specification of heart looping orientation. This is consistent with known Bmp7 stimulation of glycolysis and glycolysis suppression of cardiomyocyte differentiation. Liver/lung laterality may be specified via similar metabolic regulation of endoderm differentiation. Myo1d, found to be left-sided, was shown to regulate gut looping in mice, zebrafish, and human. Together these findings indicate metabolic regulation of left-right patterning. This could underlie high incidence of heterotaxy-related birth defects in maternal diabetes, and the association of PFKP, allosteric enzyme regulating glycolysis, with heterotaxy. This transcriptome dataset will be invaluable for interrogating birth defects involving laterality disturbance.

developmental biology↗

bioRxiv2022

Single cell transcriptomic profiling of human brain organoids reveals developmental timing- and cell-type-specific vulnerabilities induced by NRXN1 CNVs in schizophrenia

De novo mutations and copy number deletions in NRXN1 (2p16.3) pose a significant risk for schizophrenia (SCZ). It is unclear how NRXN1 deletions impact cortical development in a cell type-specific manner and disease background modulates these phenotypes. Here, we leveraged human pluripotent stem cell-derived forebrain organoid models carrying NRXN1 heterozygous deletions in isogenic and SCZ patient genetic backgrounds and conducted single-cell transcriptomic analysis over the course of brain organoid development from 3 weeks to 3.5 months. Intriguingly, while both deletions similarly impacted molecular pathways associated with ubiquitin-proteasome system, alternative splicing, and synaptic signaling in maturing glutamatergic and GABAergic neurons, SCZ-NRXN1 deletions specifically perturbed developmental trajectories of early neural progenitors and accumulated disease-specific transcriptomic signatures. Using calcium imaging, we found that both deletions led to long-lasting changes in spontaneous and synchronous neuronal networks, implicating synaptic dysfunction. Our study reveals developmental-timing- and cell-type-dependent actions of NRXN1 deletions in unique genetic contexts.

neuroscience↗

bioRxiv2022

Guided construction of single cell reference for human and mouse lung

Accurate cell type identification is a key and rate-limiting step in single cell data analysis. Single cell references with comprehensive cell types, reproducible and functional validated cell identities, and common nomenclatures are much needed by the research community to optimize automated cell type annotation and facilitate data integration, sharing, and collaboration. In the present study, we developed a novel computational pipeline to utilize the LungMAP CellCards as a dictionary to consolidate single-cell transcriptomic datasets of 104 human lungs and 17 mouse lung samples and constructed "LungMAP CellRef" and "LungMAP CellRef Seed" for both normal human and mouse lungs. "CellRef Seed" has an equivalent prediction power and produces consistent cell annotation as does "CellRef" but improves computational efficiency and simplifies its utilization for fast automated cell type annotation and online visualization. This atlas set incorporates 48 human and 40 mouse well-defined lung cell types catalogued from diverse anatomic locations and developmental time points. Using independent datasets, we demonstrated the utility of our CellRefs for automated cell type annotation analysis of both normal and disease lungs. User-friendly web interfaces were developed to support easy access and maximal utilization of the LungMAP CellRefs. LungMAP CellRefs are freely available to the pulmonary research community through fast interactive web interfaces to facilitate hypothesis generation, research discovery, and identification of cell type alterations in disease conditions.

bioinformatics↗

bioRxiv2022

CellDrift: Inferring Perturbation Responses in Temporally-Sampled Single Cell Data

Cells and tissues respond to perturbations in multiple ways that can be sensitively reflected in alterations of gene expression. Current approaches to finding and quantifying the effects of perturbations on cell-level responses over time disregard the temporal consistency of identifiable gene programs. To leverage the occurrence of these patterns for perturbation analyses, we developed CellDrift (https://github.com/KANG-BIOINFO/CellDrift), a generalized linear model-based functional data analysis method capable of identifying covarying temporal patterns of various cell types in response to perturbations. As compared to several other approaches, CellDrift demonstrated superior performance in the identification of temporally varied perturbation patterns and the ability to impute missing time points. We applied CellDrift to multiple longitudinal datasets, including COVID-19 disease progression and gastrointestinal tract development, and demonstrated its ability to identify specific gene programs associated with sequential biological processes, trajectories, and outcomes.

bioinformatics↗

bioRxiv2022

KAT5 activity regulates G0-like states in human gliomas

In solid tumors, G0-like states are likely critical for maintaining developmental hierarchies and cellular heterogeneity and promoting tumor growth/recurrence, yet little is known about tumor G0 states or regulation of their ingress/egress. To discover G0-like states and their regulators for glioblastoma (GBM), we analyzed G0 populations in an orthotopic model of GBM using single cell RNA-seq and performed a genome-wide CRISPR-Cas9 screen in patient-derived GBM stem-like cells (GSCs) for genes that trap cells in G0 when inhibited. We identify the protein acetyltransferase KAT5 as a key regulator of transcriptional, epigenetic, and proliferative heterogeneity impacting transitions into G0-like states. KAT5 activity suppresses the emergence of non-dividing subpopulations with oligodendrocyte progenitor and radial glial cell characteristics both in vitro and in a human GSC brain tumor model. In primary gliomas, KAT5 activity is dynamic with KAT5low tumor cells displaying quiescent properties, while KAT5 activity overall increases from low to high grade tumors and is associated with worse patient outcomes.

cancer biology↗

bioRxiv2022

Loss of Neurodevelopmental Gene CASK Disrupts Neural Connectivity in Human Cortical Excitatory Neurons

Loss-of-function (LOF) mutations in CASK cause severe developmental phenotypes, including microcephaly with pontine and cerebellar hypoplasia, X-linked intellectual disability, and autism. Unraveling the pathogenesis of CASK-related disorders has been challenging due to limited human cellular models to study the dynamic roles of this molecule during neuronal and synapse development. Here, we generated CASK knockout (KO) isogenic cell lines from human embryonic stem cells (hESCs) using CRISPR/Cas9 and examined gene expression, morphometrics, and synaptic function of induced neuronal cells during development. While young (immature) CASK KO neurons show robust neuronal outgrowth, mature CASK KO neurons displayed severe defects in synaptic transmission and synchronized burst activity without compromising neuronal morphology and synapse numbers. In developing human cortical neurons, CASK functioned to promote both structural integrity and establishment of cortical excitatory neuronal networks. These results lay the foundation for future studies to identify suppressors of such phenotypes relevant to human patients. HighlightsO_LICASK LOF mutations increase neuronal complexity in immature developing neurons C_LIO_LICASK LOF does not alter synapse formation and neurite complexity in mature neurons C_LIO_LISynaptic transmission and network synchronicity are compromised in CASK KO neurons C_LIO_LIDifferential gene expression analysis reveals enrichment of synaptic gene networks in mature CASK KO neurons C_LI

neuroscience↗

bioRxiv2021

LungMAP Portal Ecosystem: Systems-Level Exploration of the Lung

An improved understanding of the human lung necessitates advanced systems models informed by an ever-increasing repertoire of molecular omics, cellular, imaging and pathological datasets. To centralize and standardize information across broad lung research efforts we expanded the LungMAP.net website into a gateway portal. This portal connects a broad-spectrum of research networks, bulk and single-cell multi-omics data and a diverse collection of image data that span mammalian lung development and disease. The data are standardized across species and technologies using harmonized data and metadata models that leverage recent advances including those from the Human Cell Atlas, diverse ontologies, and the LungMAP CellCards initiative. To cultivate future discoveries, we have aggregated a diverse collection of single-cell atlases for multiple species (human, rhesus, mouse), to enable consistent queries across technologies, cohorts, age, disease and drug treatment. These atlases are provided as independent and integrated queriable datasets, with an emphasis on dynamic visualization, figure generation and reference-based classification of user-provided datasets (Azimuth). As this resource grows, we intend to increase the breadth of available interactive interfaces, data portals and datasets from LungMAP and external research efforts.

systems biology↗

bioRxiv2021

Betacoronavirus-specific alternate splicing

Viruses can subvert a number of cellular processes in order to block innate antiviral responses, and many viruses interact with cellular splicing machinery. SARS-CoV-2 infection was shown to suppress global mRNA splicing, and at least 10 SARS-CoV-2 proteins bind specifically to one or more human RNAs. Here, we investigate 17 published experimental and clinical datasets related to SARS-CoV-2 infection as well as datasets from the betacoronaviruses SARS-CoV and MERS as well as Streptococcus pneumonia, HCV, Zika virus, Dengue virus, influenza H3N2, and RSV. We show that genes showing differential alternative splicing in SARS-CoV-2 have a similar functional profile to those of SARS-CoV and MERS and affect a diverse set of genes and biological functions, including many closely related to virus biology. Additionally, the differentially spliced transcripts of cells infected by coronaviruses were more likely to undergo intron-retention, contain a pseudouridine modification and a smaller number of exons than differentially spliced transcripts in the control groups. Viral load in clinical COVID-19 samples was correlated with isoform distribution of differentially spliced genes. A significantly higher number of ribosomal genes are affected by DAS and DGE in betacoronavirus samples, and the betacoronavirus differentially spliced genes are depleted for binding sites of RNA-binding proteins. Our results demonstrate characteristic patterns of differential splicing in cells infected by SARS-CoV-2, SARS-CoV, and MERS, potentially modifying a broad range of cellular functions and affecting a diverse set of genes and biological functions.

genomics↗

bioRxiv2021

A Web Portal and Workbench for Biological Dissection of Single Cell COVID-19 Host Responses

Numerous studies have provided single-cell transcriptome profiles of host responses to SARS-CoV-2 infection. Critically lacking however is a datamine that allows users to compare and explore cell profiles to gain insights and develop new hypotheses. To accomplish this, we harmonized datasets from COVID-19 and other control condition blood, bronchoalveolar lavage, and tissue samples, and derived a compendium of gene signature modules per cell type, subtype, clinical condition, and compartment. We demonstrate approaches to probe these via a new interactive web portal (http://toppcell.cchmc.org/ COVID-19). As examples, we develop three hypotheses: (1) a multicellular signaling cascade among alternatively differentiated monocyte-derived macrophages whose tasks include T cell recruitment and activation; (2) novel platelet subtypes with drastically modulated expression of genes responsible for adhesion, coagulation and thrombosis; and (3) a multilineage cell activator network able to drive extrafollicular B maturation via an ensemble of genes strongly associated with risk for developing post-viral autoimmunity.

immunology↗

bioRxiv2021

iPSC modeling shows uncompensated mitochondrial mediated oxidative stress underlies early heart failure in hypoplastic left heart syndrome

Hypoplastic left heart syndrome (HLHS) is a severe congenital heart defect with 30% mortality from heart failure (HF) in the first year of life, but why only some patients suffer early-HF and its cause remain unknown. Modeling using induced pluripotent stem cell-derived cardiomyocytes (iPSC-CM) showed early-HF patient iPSC-CM have increased apoptosis, redox stress, and failed antioxidant response. This was associated with mitochondrial permeability transition pore (mPTP) opening, mitochondrial hyperfusion and respiration defects. Whereas iPSC-CM from patients without early-HF had hyper-elevated antioxidant response with increased mitochondrial fission and mitophagy. Single cell transcriptomics showed dichotomization by HF outcome, with mitochondrial dysfunction and endoplasmic reticulum (ER) stress associated with early-HF. Importantly, oxidative stress and apoptosis associated with early HF were rescued by sildenafil inhibition of mPTP opening or TUDCA suppression of ER stress. Together these findings demonstrate a new paradigm for modeling clinical outcome in iPSC-CM, demonstrating uncompensated mitochondrial oxidative stress underlies early HF in HLHS.

cell biology↗

bioRxiv2021

Stratification of Risk of Progression to Colectomy in Ulcerative Colitis using Measured and Predicted Gene Expression

An important goal of clinical genomics is to be able to estimate the risk of adverse disease outcomes. Between 5% and 10% of ulcerative colitis (UC) patients require colectomy within five years of diagnosis, but polygenic risk scores (PRS) utilizing findings from GWAS are unable to provide meaningful prediction of this adverse status. By contrast, in Crohns disease, gene expression profiling of GWAS-significant genes does provide some stratification of risk of progression to complicated disease in the form of a Transcriptional Risk Score (TRS). Here we demonstrate that both measured (TRS) and polygenic predicted gene expression (PPTRS) identify UC patients at 5-fold elevated risk of colectomy with data from the PROTECT clinical trial and UK Biobank population cohort studies, independently replicated in an NIDDK-IBDGC dataset. Prediction of gene expression from relatively small transcriptome datasets can thus be used in conjunction with transcriptome-wide association studies to stratify risk of disease complications.

genetics↗

bioRxiv2020

Cross-Platform Validation of Neurotransmitter Release Impairments in Schizophrenia Patient-Derived NRXN1-Mutant Neurons

Heterozygous NRXN1 deletions constitute the most prevalent currently known single-gene mutation predisposing to schizophrenia. Previous studies showed that engineered heterozygous NRXN1 deletions impaired neurotransmitter release in human neurons, suggesting a synaptic pathophysiological mechanism. Utilizing this observation for drug discovery, however, requires confidence in its robustness and validity. Here, we describe a multi-center effort to test the generality of this pivotal observation, using independent analyses at two laboratories of patient-derived and newly engineered human neurons with heterozygous NRXN1 deletions. We show that in neurons that were trans-differentiated from induced pluripotent stem cells derived from three NRXN1-deletion patients, the same impairment in neurotransmitter release was observed as in engineered NRXN1-deficient neurons. This impairment manifested as a decrease in spontaneous synaptic events and in evoked synaptic responses, and an alteration in synaptic paired-pulse depression. Nrxn1-deficient mouse neurons generated from embryonic stem cells by the same method as human neurons did not exhibit impaired neurotransmitter release, suggesting a human-specific phenotype. NRXN1 deletions produced a reproducible increase in the levels of CASK, an intracellular NRXN1-binding protein, and were associated with characteristic gene expression changes. Thus, heterozygous NRXN1 deletions robustly impair synaptic function in human neurons regardless of genetic background, enabling future drug discovery efforts.

neuroscience↗