Biology subjects

Regev, A.

Publications and source records attributed to Regev, A..

At least 37 records · Page 2Linked to original sources

bioRxiv2025

CellUntangler: separating distinct biological signals in single-cell data with deep generative models

Single-cell RNA-seq data have provided new insights into intracellular and intercellular processes. Because multiple processes are active in each cell simultaneously, such as its cell type program, differentiation, the cell cycle, and environmental responses, their respective signals can confound one another, requiring methods that can separate and filter different complex biological signals. Each such signal is based on different gene activities and can define different relationships between cells. However, existing methods often focus on a single process or rely on overly restrictive assumptions, thus removing, rather than disentangling biological signals. Here, we develop CellUntangler, a deep generative model that embeds cells into a flexible latent space composed of multiple subspaces, each designed with an appropriate geometry to capture a distinct signal. We apply CellUntangler to datasets containing only cycling cells and both cycling and non-cycling cells, generating embeddings in which the cell cycle signal is disentangled from non-cell cycle specific signals, such as cell type or differentiation trajectory. We demonstrate CellUntanglers extensibility by using it to capture and separate spatial from non-spatial signals. With CellUntangler, we can obtain latent embeddings that capture various biological signals and perform enhancement or filtering at the gene expression level for downstream analyses.

bioinformatics↗

bioRxiv2024

PathOmCLIP: Connecting tumor histology with spatial gene expression via locally enhanced contrastive learning of Pathology and Single-cell foundation model

Tumor morphological features from histology images are a cornerstone of clinical pathology, diagnostic biomarkers, and basic cancer biology research. Spatial transcriptomics, which provides spatially resolved gene expression profiles overlaid on histology images, offers a unique opportunity to integrate morphological and expression features, thereby deepening our understanding of tumor biology. However, spatial transcriptomics experiments with patient samples in either clinical trials or clinical care are costly and challenging, whereas histology images are generated routinely and available for many legacy prospective cohorts of disease progression and outcomes in well-annotated cohorts. Inferring spatial transcriptomics profiles computationally from these histology images would significantly expand our understanding of tumor biology, but paired data for training multi-modal spatial-histology models remains limited. Here, we tackle this challenge by incorporating performant foundation models pre-trained on massive datasets of pathology images and single-cell RNA-Seq, respectively, which provide useful embeddings to underpin multi-modal models. To this end, we developed PathOmCLIP, a model trained with contrastive loss to create a joint-embedding space between a histopathology foundation model and a single-cell RNA-seq foundation model. We incorporate a set transformer to gather localized neighborhood tumor architecture following contrastive training, which further enhances performance and is necessary to obtain robust results. We validate PathOmCLIP across five tumor types and achieve significant performance improvements in gene expression prediction tasks over other methods. PathOmCLIP can be applied to many archived histology images, unlocking valuable clinical information and facilitating new biomarker discoveries.

bioinformatics↗

bioRxiv2024

Learning multi-cellular representations of single-cell transcriptomics data enables characterization of patient-level disease states

Single-cell RNA-seq (scRNA-seq) has become a prominent tool for studying human biology and disease. The availability of massive scRNA-seq datasets and advanced machine learning techniques has recently driven the development of single-cell foundation models that provide informative and versatile cell representations based on expression profiles. However, to understand disease states, we need to consider entire tissue ecosystems, simultaneously considering many different interacting cells. Here, we tackle this challenge by generating patient-level representations derived from multi-cellular expression context measured with scRNA-seq of tissues. We develop PaSCient, a novel model that employs a multi-level representation learning paradigm and provides importance scores at the individual cell and gene levels for fine-grained analysis across multiple cell types and gene programs characteristic of a given disease. We apply PaSCient to learn a disease model across a large-scale scRNA-seq atlas of 24.3 million cells from over 5,000 patients. Comprehensive and rigorous benchmarking demonstrates the superiority of PaSCient in disease classification and its multiple downstream applications, including dimensionality reduction, gene/cell type prioritization, and patient subgroup discovery.

bioinformatics↗

bioRxiv2024

A high-throughput phenotypic screen combined with an ultra-large-scale deep learning-based virtual screening reveals novel scaffolds of antibacterial compounds

The proliferation of multi-drug-resistant bacteria underscores an urgent need for novel antibiotics. Traditional discovery methods face challenges due to limited chemical diversity, high costs, and difficulties in identifying structurally novel compounds. Here, we explore the integration of small molecule high-throughput screening with a deep learning-based virtual screening approach to uncover new antibacterial compounds. Leveraging a diverse library of nearly 2 million small molecules, we conducted comprehensive phenotypic screening against a sensitized Escherichia coli strain that, at a low hit rate, yielded thousands of hits. We trained a deep learning model, GNEprop, to predict antibacterial activity, ensuring robustness through out-of-distribution generalization techniques. Virtual screening of over 1.4 billion compounds identified potential candidates, of which 82 exhibited antibacterial activity, illustrating a 90X improved hit rate over the high-throughput screening experiment GNEprop was trained on. Importantly, a significant portion of these newly identified compounds exhibited high dissimilarity to known antibiotics, indicating promising avenues for further exploration in antibiotic discovery.

bioinformatics↗

bioRxiv2024

LAT encodes T cell activation pathway balance

Immune cells transduce environmental stimuli into responses essential for host health via complex signaling cascades. T cells, in particular, leverage their unique T cell receptors (TCRs) to detect specific Human Leukocyte Antigen (HLA)-presented peptides. TCR activation is then relayed via linker for activation of T cells (LAT), a TCR-proximal disordered adapter protein, which organizes protein partners and mediates the propagation of signals down diverse pathways including NFAT and AP-1. Here, we studied how balanced downstream pathway activation is encoded in the amino acid sequence of LAT. To comprehensively profile the sequence-function relationship of LAT, we developed a pooled, single-cell, high-content screening approach in which a large series of mutants in the LAT protein were analyzed to characterize their effects on T cell activation. Measuring epigenetic, transcriptomic, and cell surface protein dynamics of single cells harboring distinct LAT mutants, we found functional regions spanning over 40% of the LAT amino acid sequence. Conserved sequence motifs for protein interactions along with charge distribution are critical sequence features, and contribute to interpretation of human genetic variation in LAT. While mutant defect severity spans from moderate to complete loss of function, nearly all defective mutants, irrespective of their position in LAT, confer balanced defects across all downstream pathways. To understand the molecular basis for this observation, we performed proximal protein labeling which demonstrated that disruption of LAT interaction with a single partner protein indirectly disrupts other partner interactions, likely through the dual roles of these proteins as effectors of downstream pathways and bridging factors between LAT molecules. Overall, we report widely distributed functional regions throughout a disordered adapter and a precise physical organization of LAT and interacting molecules which constrains signaling outputs. More broadly, we describe an approach for interrogating sequence-function relationships for proteins with complex activities across regulatory layers of the cell.

genomics↗

bioRxiv2024

Experimental and Computational Methods for Allelic Imbalance Analysis from Single-Nucleus RNA-seq Data

Single-cell RNA-seq (scRNA-seq) is emerging as a powerful tool for understanding gene function across diverse cells. Recently, this has included the use of allele-specific expression (ASE) analysis to better understand how variation in the human genome affects RNA expression at the single-cell level. We reasoned that because intronic reads are more prevalent in single-nucleus RNA-Seq (snRNA-Seq), and introns are under lower purifying selection and thus enriched for genetic variants, that snRNA-seq should facilitate single-cell analysis of ASE. Here we demonstrate how experimental and computational choices can improve the results of allelic imbalance analysis. We explore how experimental choices, such as RNA source, read length, sequencing depth, genotyping, etc., impact the power of ASE-based methods. We developed a new suite of computational tools to process and analyze scRNA-seq and snRNA-seq for ASE. As hypothesized, we extracted more ASE information from reads in intronic regions than those in exonic regions and show how read length can be set to increase power. Additionally, hybrid selection improved our power to detect allelic imbalance in genes of interest. We also explored methods to recover allele-specific isoform expression levels from both long- and short-read snRNA-seq. To further investigate ASE in the context of human disease, we applied our methods to a Parkinsons disease cohort of 94 individuals and show that ASE analysis had more power than eQTL analysis to identify significant SNP/gene pairs in our direct comparison of the two methods. Overall, we provide an end-to-end experimental and computational approach for future studies.

genomics↗

bioRxiv2024

Tissue and cellular spatiotemporal dynamics in colon aging

Tissue structure and molecular circuitry in the colon can be profoundly impacted by systemic age-related effects, but many of the underlying molecular cues remain unclear. Here, we built a cellular and spatial atlas of the colon across three anatomical regions and 11 age groups, encompassing [~]1,500 mouse gut tissues profiled by spatial transcriptomics and [~]400,000 single nucleus RNA-seq profiles. We developed a new computational framework, cSplotch, which learns a hierarchical Bayesian model of spatially resolved cellular expression associated with age, tissue region, and sex, by leveraging histological features to share information across tissue samples and data modalities. Using this model, we identified cellular and molecular gradients along the adult colonic tract and across the main crypt axis, and multicellular programs associated with aging in the large intestine. Our multi-modal framework for the investigation of cell and tissue organization can aid in the understanding of cellular roles in tissue-level pathology.

systems biology↗

bioRxiv2024

Natural language processing charts transcriptome evolution to design combination cancer therapies

Combination treatment-the mainstay of cancer therapy-often fails because treatment interactions evoke complex resistance mechanisms that are hard to predict. Designing combination therapy to prevent treatment resistance is especially challenging for rare cancers. Here, we introduce REsistance through COntext DRift (RECODR): a computational pipeline that tracks how genes change their transcriptome context across cancer development and drug treatment conditions. By applying RECODR to a genetically modified mouse model of choroid plexus carcinoma-a rare brain tumour of young children-we identified patterns of transcriptome evolution, cellular heterogeneity and treatment targets that emerged as tumours were initiated and resisted combination treatment. This enabled the prediction of treatment resistance mechanisms and the design of highly effective therapeutic protocols that avoided treatment failure. RECODR can describe complex and dynamic changes in normal and diseased tissues that could be applied to multimodal data from a variety of settings, mitigating treatment resistance across cancers and other diseases.

cancer biology↗

bioRxiv2024

Identification of markers for the isolation of neuron-specific extracellular vesicles

Extracellular vesicles (EVs) are released by all cells and contain RNA and protein from their cell of origin. EVs in biofluids could be used as diagnostic biomarkers to non-invasively report the state of inaccessible cells, such as neurons in the brain. As biofluids such as cerebrospinal fluid (CSF) and plasma contain EVs originating from many different cells, isolating cell type-specific EVs and measuring their cargo could help determine the state of specific cell types. Here, we demonstrate an approach aiming to immuno-isolate EVs from neurons based on neuron-derived protein surface markers. We first developed a framework to select transmembrane proteins suitable as neuron-specific EV markers based on gene expression and EV proteomics data. Leveraging a novel, high-purity EV isolation method we developed, we further cataloged the proteins present on EVs in human CSF and plasma. Using ultrasensitive immunoassays against several of the predicted neuron-specific proteins, we confirmed one marker, NRXN3 as present on EVs in CSF and plasma by size exclusion chromatography (SEC) and density gradient centrifugation (DGC). Finally, we developed efficient EV immuno-isolation methods and applied them to isolate NRXN3+ EVs. Our study provides a general methodology for the isolation of cell-type specific EVs and paves the way for the use of neuron-derived EVs to study and diagnose neurological disease.

molecular biology↗

bioRxiv2024

Metric Mirages in Cell Embeddings

Although biological studies increasingly rely on embeddings of single cell profiles, the quality of these embeddings can be challenging to assess. Such evaluations are especially important for avoiding misleading biological interpretations, assessing the accuracy of integration methods, and establishing the zero-shot capabilities of foundational models. Here, we posit that current evaluation metrics can be highly misleading. We show this by training a three-layer perceptron, Islander , which outperforms all 11 leading embedding methods on a diverse set of cell atlases, but in fact distorts biological structures, limiting its utility for biological discovery. We then present a metric, scGraph, to flag such distortions. Our work should help learn more robust and reliable cell embeddings.

bioinformatics↗

bioRxiv2024

A Supervised Contrastive Framework for Learning Disentangled Representations of Cell Perturbation Data

CRISPR technology, combined with single-cell RNA-Seq, has opened the way to large scale pooled perturbation screens, allowing more systematic interrogations of gene functions in cells at scale. However, such Perturb-seq data poses many analysis challenges, due to its high-dimensionality, high level of technical noise, and variable Cas9 efficiency. The single-cell nature of the data also poses its own challenges, as we observe the heterogeneity of phenotypes in the unperturbed cells, along with the effect of the perturbations. All in all, these characteristics make it difficult to discern subtler effects. Existing tools, like mixscape and ContrastiveVI, provide partial solutions, but may oversimplify biological dynamics, or have low power to characterize perturbations with a smaller effect size. Here, we address these limitations by introducing the Supervised Contrastive Variational Autoencoder (SC-VAE). SC-VAE integrates guide RNA identity with gene expression data, ensuring a more discriminative analysis, and adopts the Hilbert-Schmidt Independence Criterion as a way to achieve disentangled representations, separating the heterogeneity in the control population from the effect of the perturbations. Evaluation on large-scale data sets highlights SC-VAEs superior sensitivity in identifying perturbation effects compared to ContrastiveVI, scVI and PCA. The perturbation embeddings better reflect known protein complexes (evaluated on CORUM), while its classifier offers promise in identifying assignment errors and cells escaping the perturbation phenotype. SC-VAE is readily applicable across diverse perturbation data sets.

bioinformatics↗

bioRxiv2023

Highly multiplexed, image-based pooled screens in primary cells and tissues with PerturbView

Optical pooled screening (OPS) is a highly scalable method for linking image-based phenotypes with cellular perturbations. However, it has thus far been restricted to relatively low-plex phenotypic readouts in cancer cell lines in culture, due to limitations associated with in situ sequencing (ISS) of perturbation barcodes. Here, we developed PerturbView, an OPS technology that leverages in vitro transcription (IVT) to amplify barcodes prior to ISS, enabling screens with highly multiplexed phenotypic readouts across diverse systems, including primary cells and tissues. We demonstrate PerturbView in iPSC-derived neurons, primary immune cells, and tumor tissue sections from animal models. In a screen of immune signaling pathways in primary bone marrow-derived macrophages, PerturbView uncovered both known and novel regulators of NF{kappa}B signaling. Furthermore, we combined PerturbView with spatial transcriptomics in tissue sections from a mouse xenograft model, paving the way to in vivo screens with rich optical and transcriptomic phenotypes. PerturbView broadens the scope of OPS to a wide range of models and applications.

systems biology↗

bioRxiv2023

Sequential Optimal Experimental Design of Perturbation Screens Guided by Multi-modal Priors

Understanding a cells expression response to genetic perturbations helps to address important challenges in biology and medicine, including the function of gene circuits, discovery of therapeutic targets and cell reprogramming and engineering. In recent years, Perturb-seq, pooled genetic screens with single cell RNA-seq (scRNA-seq) readouts, has emerged as a common method to collect such data. However, irrespective of technological advances, because combinations of gene perturbations can have unpredictable, non-additive effects, the number of experimental configurations far exceeds experimental capacity, and for certain cases, the number of available cells. While recent machine learning models, trained on existing Perturb-seq data sets, can predict perturbation outcomes with some degree of accuracy, they are currently limited by sub-optimal training set selection and the small number of cell contexts of training data, leading to poor predictions for unexplored parts of perturbation space. As biologists deploy Perturb-seq across diverse biological systems, there is an enormous need for algorithms to guide iterative experiments while exploring the large space of possible perturbations and their combinations. Here, we propose a sequential approach for designing Perturb-seq experiments that uses the model to strategically select the most informative perturbations at each step for subsequent experiments. This enables a significantly more efficient exploration of the perturbation space, while predicting the effect of the rest of the unseen perturbations with high-fidelity. Analysis of a previous large-scale Perturb-seq experiment reveals that our setting is severely restricted by the number of examples and rounds, falling into a non-conventional active learning regime called "active learning on a budget". Motivated by this insight, we develop IO_SCPLOWTERC_SCPLOWPO_SCPLOWERTC_SCPLOW, a novel active learning method that exploits rich and multi-modal prior knowledge in order to efficiently guide the selection of subsequent perturbations. Using prior knowledge for this task is novel, and crucial for successful active learning on a budget. We validate IO_SCPLOWTERC_SCPLOWPO_SCPLOWERTC_SCPLOW using insilico benchmarking of active learning, constructed from a large-scale CRISPRi Perturb-seq data set. We find that IO_SCPLOWTERC_SCPLOWPO_SCPLOWERTC_SCPLOW outperforms other active learning strategies by reaching comparable accuracy at only a third of the number of perturbations profiled as the next best method. Overall, our results demonstrate the potential of sequentially designing perturbation screens through IO_SCPLOWTERC_SCPLOWPO_SCPLOWERTC_SCPLOW.

bioinformatics↗

bioRxiv2023

A deep lung cell atlas reveals cytokine-mediated lineage switching of a rare cell progenitor of the human airway epithelium

The human airway contains specialized rare epithelial cells whose roles in respiratory disease are not well understood. Ionocytes express the Cystic Fibrosis Transmembrane Conductance Regulator (CFTR), while chemosensory tuft cells express asthma-associated alarmins. However, surprisingly, exceedingly few mature tuft cells have been identified in human lung cell atlases despite the ready identification of rare ionocytes and neuroendocrine cells. To identify human rare cell progenitors and define their lineage relationship to mature tuft cells, we generated a deep lung cell atlas containing 311,748 single cell RNA-Seq (scRNA-seq) profiles from discrete anatomic sites along the large and small airways and lung lobes of explanted donor lungs that could not be used for organ transplantation. Of 154,222 airway epithelial cells, we identified 687 ionocytes (0.45%) that are present in similar proportions in both large and small airways, suggesting that they may contribute to both large and small airways pathologies in CF. In stark contrast, we recovered only 3 mature tuft cells (0.002%). Instead, we identified rare bipotent progenitor cells that can give rise to both ionocytes and tuft cells, which we termed tuft-ionocyte progenitor cells (TIP cells). Remarkably, the cycling fraction of these TIP cells was comparable to that of basal stem cells. We used scRNA-seq and scATAC-seq to predict transcription factors that mark this novel rare cell progenitor population and define intermediate states during TIP cell lineage transitions en route to the differentiation of mature ionocytes and tuft cells. The default lineage of TIP cell descendants is skewed towards ionocytes, explaining the paucity of mature tuft cells in the human airway. However, Type 2 and Type 17 cytokines, associated with asthma and CF, diverted the lineage of TIP cell descendants in vitro, resulting in the differentiation of mature tuft cells at the expense of ionocytes. Consistent with this model of mature tuft cell differentiation, we identify mature tuft cells in a patient who died from an asthma flare. Overall, our findings suggest that the immune signaling pathways active in asthma and CF may skew the composition of disease-relevant rare cells and illustrate how deep atlases are required for identifying physiologically-relevant scarce cell populations.

cell biology↗

bioRxiv2023

Multi-ContrastiveVAE disentangles perturbation effects in single cell images from optical pooled screens

Optical pooled screens (OPS) enable comprehensive and cost-effective interrogation of gene function by measuring microscopy images of millions of cells across thousands of perturbations. However, the analysis of OPS data still mainly relies on hand-crafted features, even though these are difficult to deploy across complex data sets. This is because most unsupervised feature extraction methods based on neural networks (such as auto-encoders) have difficulty isolating the effect of perturbations from the natural variations across cells and experimental batches. Here, we propose a contrastive analysis framework that can more effectively disentangle the phenotypes caused by perturbation from natural cell-cell heterogeneity present in an unperturbed cell population. We demonstrate this approach by analyzing a large data set of over 30 million cells imaged across more than 5, 000 genetic perturbations, showing that our method significantly outperforms traditional approaches in generating biologically-informative embeddings and mitigating technical artifacts. Furthermore, the interpretable part of our model distinguishes perturbations that generate novel phenotypes from the ones that only shift the distribution of existing phenotypes. Our approach can be readily applied to other small-molecule and genetic perturbation data sets with highly multiplexed images, enhancing the efficiency and precision in identifying and interpreting perturbation-specific phenotypic patterns, paving the way for deeper insights and discoveries in OPS analysis.

bioinformatics↗

bioRxiv2023

Multi-modal skin atlas associates a multicellular immune-stromal community with altered cornification and T cell expansion in atopic dermatitis

In healthy skin, a cutaneous immune system maintains the balance between tolerance towards innocuous environmental antigens and immune responses against pathological agents. In atopic dermatitis (AD), barrier and immune dysfunction result in chronic tissue inflammation. Our understanding of the skin tissue ecosystem in AD remains incomplete with regard to the hallmarks of pathological barrier formation, and cellular state and clonal composition of disease-promoting cells. Here, we generated a multi-modal cell census of 310,691 cells spanning 86 cell subsets from whole skin tissue of 19 adult individuals, including non-lesional and lesional skin from 11 AD patients, and integrated it with 396,321 cells from four studies into a comprehensive human skin cell atlas in health and disease. Reconstruction of human keratinocyte differentiation from basal to cornified layers revealed a disrupted cornification trajectory in AD. This disrupted epithelial differentiation was associated with signals from a unique immune and stromal multicellular community comprised of MMP12+ dendritic cells (DCs), mature migratory DCs, cycling ILCs, NK cells, inflammatory CCL19+ IL4I1+ fibroblasts, and clonally expanded IL13+IL22+IL26+ T cells with overlapping type 2 and type 17 characteristics. Cell subsets within this immune and stromal multicellular community were connected by multiple inter-cellular positive feedback loops predicted to impact community assembly and maintenance. AD GWAS gene expression was enriched both in disrupted cornified keratinocytes and in cell subsets from the lesional immune and stromal multicellular community including IL13+IL22+IL26+ T cells and ILCs, suggesting that epithelial or immune dysfunction in the context of the observed cellular communication network can initiate and then converge towards AD. Our work highlights specific, disease-associated cell subsets and interactions as potential targets in progression and resolution of chronic inflammation.

molecular biology↗

bioRxiv2023

Multi-donor human cortical Chimeroids reveal individual susceptibility to neurotoxic triggers

Inter-individual genetic variation affects the susceptibility to and progression of many diseases. Efforts to study the molecular mechanisms mediating the impact of human genetic variation on normal development and disease phenotypes are limited, however, by the paucity of faithful cellular human models, and the difficulty of scaling current systems to represent multiple individuals. Here, we present human brain "Chimeroids", a highly reproducible, multi-donor human brain cortical organoid model generated by the co-development of cells from a panel of individual donors in a single organoid, while maintaining fidelity to endogenous tissue. By reaggregating cells from multiple single-donor organoids at the neural stem or committed progenitor cell stage, we generate Chimeroids in which each donor produces all cell lineages of the cerebral cortex, even when using pluripotent stem cell lines with notable growth biases. We leveraged Chimeroids to investigate inter-individual variation in susceptibility to neurotoxic stressors that exhibit high clinical phenotypic variability: ethanol and the anti-epileptic drug valproic acid. Individual donors varied in both the penetrance of the effect on target cell types, and the molecular phenotype within each affected cell type. Our results show that human genetic background may be an important mediator of neurotoxin susceptibility and introduce Chimeroids as a scalable system for high-throughput investigation of the contribution of human genetic variation to brain development and disease.

neuroscience↗

bioRxiv2023

Scalable querying of human cell atlases via a foundational model reveals commonalities across fibrosis-associated macrophages

Single-cell RNA-seq (scRNA-seq) studies have profiled over 100 million human cells across diseases, developmental stages, and perturbations to date. A singular view of this vast and growing expression landscape could help reveal novel associations between cell states and diseases, discover cell states in unexpected tissue contexts, and relate in vivo cells to in vitro models. However, these require a common, scalable representation of cell profiles from across the body, a general measure of their similarity, and an efficient way to query these data. Here, we present SCimilarity, a metric learning framework to learn and search a unified and interpretable representation that annotates cell types and instantaneously queries for a cell state across tens of millions of profiles. We demonstrate SCimilarity on a 22.7 million cell corpus assembled across 399 published scRNA-seq studies, showing accurate integration, annotation and querying. We experimentally validated SCimilarity by querying across tissues for a macrophage subset originally identified in interstitial lung disease, and showing that cells with similar profiles are found in other fibrotic diseases, tissues, and a 3D hydrogel system, which we then repurposed to yield this cell state in vitro. SCimilarity serves as a foundational model for single cell gene expression data and enables researchers to query for similar cellular states across the entire human body, providing a powerful tool for generating novel biological insights from the growing Human Cell Atlas.

bioinformatics↗