Biology subjects

Azizi, E.

Publications and source records attributed to Azizi, E..

30 records · Page 2Linked to original sources

bioRxiv2023

DIISCO: A Bayesian framework for inferring dynamic intercellular interactions from time-series single-cell data

Characterizing cell-cell communication and tracking its variability over time is essential for understanding the coordination of biological processes mediating normal development, progression of disease, or responses to perturbations such as therapies. Existing tools lack the ability to capture time-dependent intercellular interactions, such as those influenced by therapy, and primarily rely on existing databases compiled from limited contexts. We present DIISCO, a Bayesian framework for characterizing the temporal dynamics of cellular interactions using single-cell RNA-sequencing data from multiple time points. Our method uses structured Gaussian process regression to unveil time-resolved interactions among diverse cell types according to their co-evolution and incorporates prior knowledge of receptor-ligand complexes. We show the interpretability of DIISCO in simulated data and new data collected from CAR-T cells co-cultured with lymphoma cells, demonstrating its potential to uncover dynamic cell-cell crosstalk. AvailabilityDIISCO is publicly accessible at https://github.com/azizilab/DIISCO_public. All data will be deposited to GEO upon publication.

bioinformatics↗

bioRxiv2023

Deep generative model deciphers derailed trajectories in acute myeloid leukemia

Biological insights often depend on comparing conditions such as disease and health, yet we lack effective computational tools for integrating single-cell genomics data across conditions or characterizing transitions from normal to deviant cell states. Here, we present Decipher, a deep generative model that characterizes derailed cell-state trajectories. Decipher jointly models and visualizes gene expression and cell state from normal and perturbed single-cell RNA-seq data, revealing shared and disrupted dynamics. We demonstrate its superior performance across diverse contexts, including in pancreatitis with oncogene mutation, acute myeloid leukemia, and gastric cancer.

bioinformatics↗

bioRxiv2023

CellStitch: 3D Cellular Anisotropic Image Segmentation via Optimal Transport

BackgroundSpatial mapping of transcriptional states provides valuable biological insights into cellular functions and interactions in the context of the tissue. Accurate 3D cell segmentation is a critical step in the analysis of this data towards understanding diseases and normal development in situ. Current approaches designed to automate 3D segmentation include stitching masks along one dimension, training a 3D neural network architecture from scratch, and reconstructing a 3D volume from 2D segmentations on all dimensions. However, the applicability of existing methods is hampered by inaccurate segmentations along the non-stitching dimensions, the lack of high-quality diverse 3D training data, and inhomogeneity among different dimensions; as a result, they have not been widely used in practice. MethodsTo address these challenges, we formulate the problem of finding cell correspondence across layers with a novel optimal transport (OT) approach. We propose CellStitch, a flexible pipeline that segments cells from 3D images without requiring large amounts of 3D training data. We further extend our method to interpolate internal slices from highly anisotropic cell images to recover isotropic cell morphology. ResultsWe evaluated the performance of CellStitch through eight 3D plant microscopic datasets with diverse anisotropic levels and cell shapes. CellStitch substantially outperforms the state-of-the art methods on anisotropic images, and achieves comparable segmentation quality against competing methods in isotropic setting. We benchmarked and reported 3D segmentation results of all the methods with instance-level precision, recall and average precision (AP) metrics. ConclusionThe proposed OT-based 3D segmentation pipeline outperformed the existing state-of-the-art methods on different datasets with nonzero anisotropy, providing high fidelity recovery of 3D cell morphology from microscopic images.

bioinformatics↗

bioRxiv2022

Starfysh reveals heterogeneous spatial dynamics in the breast tumor microenvironment

Spatially-resolved gene expression profiling provides valuable insight into tissue organization and cell-cell crosstalk; however, spatial transcriptomics (ST) lacks single-cell resolution. Current ST analysis methods require single-cell RNA sequencing data as a reference for a rigorous interpretation of cell states and do not utilize associated histology images. Significant sample variation further complicates the integration of ST datasets, which is essential for identifying commonalities across tissues or altered cellular wiring in disease. Here, we present Starfysh, the first comprehensive computational toolbox for joint modeling of ST and histology data, dissection of refined cell states, and systematic integration of multiple ST datasets from complex tissues. Starfysh uses an auxiliary deep generative model that incorporates archetypal analysis and any known cell state markers to avoid the need for a single-cell-resolution reference in characterizing known or novel tissue-specific cell states. Additionally, Starfysh improves the characterization of spatial dynamics in complex tissues by leveraging histology images and enables the comparison of niches as spatial "hubs" across tissues. Integrative analysis of primary estrogen receptor-positive (ER+) breast cancer, triple-negative breast cancer (TNBC), and metaplastic breast cancer (MBC) tumors using Starfysh led to the identification of heterogeneous patient- and disease-specific hubs as well as a shared stromal hub with varying spatial orientation. Our results show the ability to delineate the spatial co-evolution of tumor and immune cell states and their crosstalk underlying intratumoral heterogeneity in TNBC and revealed metabolic reprogramming shaping immunosuppressive hubs in aggressive MBC. Starfysh is publicly available (https://github.com/azizilab/starfysh).

bioinformatics↗

bioRxiv2022

Deep generative modeling for quantifying sample-level heterogeneity in single-cell omics

The field of single-cell genomics is now observing a marked increase in the prevalence of cohort-level studies that include hundreds of samples and feature complex designs. These data have tremendous potential for discovering how sample or tissue-level phenotypes relate to cellular and molecular composition. However, current analyses are based on simplified representations of these data by averaging information across cells. We present MrVI, a deep generative model designed to realize the potential of cohort studies at the single-cell level. MrVI tackles two fundamental and intertwined problems: stratifying samples into groups and evaluating the cellular and molecular differences between groups, both without requiring a priori grouping of cells into types or states. Due to its single-cell perspective, MrVI is able to detect clinically relevant stratifications of patients in COVID-19 and inflammatory bowel disease (IBD) cohorts that are only manifested in certain cellular subsets, thus enabling new discoveries that would otherwise be overlooked. Similarly, we demonstrate that MrVI can de-novo identify groups of small molecules with similar biochemical properties and evaluate their effects on cellular composition and gene expression in large-scale perturbation studies. MrVI is available as open source at scvi-tools.org.

bioinformatics↗

bioRxiv2022

KINOMO: A non-negative matrix factorization framework for recovering intra- and inter-tumoral heterogeneity from single-cell RNA-seq data

Single-cell RNA-sequencing (scRNA-seq) is a powerful technology to uncover cellular heterogeneity in tumor ecosystems. Due to differences in underlying gene load, direct comparison between patient samples is challenging, and this is further complicated by the sparsity of data matrices in scRNA-seq. Here, we present a factorization method called KINOMO (Kernel dIfferentiability correlation-based NOn-negative Matrix factorization algorithm using Kullback-Leibler divergence loss Optimization). This tool uses quadratic approximation approach for error correction and an iterative multiplicative approach, which improves the quality assessment of NMF-identified factorization, while mitigating biases introduced by inter-patient genomic variability. We benchmarked this new approach against nine different methods across 15 scRNA-seq experiments and find that KINOMO outperforms prior methods when evaluated with an adjusted Rand index (ARI), ranging 0.82-0.91 compared to 0.68-0.77. Thus, KINOMO provides an improved approach for determining coherent transcriptional programs (and meta-programs) from scRNA-seq data of cancer tissues, enabling comparison of patients with variable genomic backgrounds. ClassificationPhysical Sciences (Applied Mathematics; Biophysics and Computational Biology), Biological Sciences (Applied Biological Sciences; Biophysics and Computational Biology; Medical Sciences; Systems Biology.). Significance StatementIdentification of shared or distinct cell programs in single-cell RNA-seq data of patient cancer cells is challenging due to underlying variability of gene load which determines transcriptional output. We developed an analytical approach to define transcriptional variability more accurately across patients and therefore enable comparison of program expression despite inherent genetic heterogeneity. Thus, this method overcomes challenges not adequately addressed by other methods broadly used for the analysis of single-cell genomics data.

systems biology↗

bioRxiv2022

Multi-modal single-cell and whole-genome sequencing of minute, frozen specimens to propel clinical applications

Single-cell genomics are enabling technologies, but their broad clinical application remains challenging. We report an easily adaptable approach for single-cell transcriptome and T cell receptor (TCR)-sequencing, and matched whole-genome sequencing from tiny, frozen clinical specimens. We achieve similar quality and biological outputs while reducing artifactual signals compared to data from matched fresh tissue samples. Profiling sequentially collected melanoma samples from the KEYNOTE-001 trial, we resolve cellular, genomic, and clonotype dynamics that encapsulate molecular patterns of tumor evolution during anti-PD-1 therapy. To demonstrate applicability to banked biospecimens of rare diseases, we generate a large uveal melanoma liver metastasis single-cell and matched WGS atlas, which revealed niche-specific impairment of clonal T cell expansion. This study provides a foundational framework for propelling single-cell genomics to the clinical arena.

genomics↗

bioRxiv2021

Mapping morphological malformation to genetic dysfunction in blood vessel organoids with 22q11.2 Deletion Syndrome

DiGeorge Syndrome, or 22q11.2 deletion syndrome (22q11.2 DS), is a genetic disorder caused by microdeletions in chromosome 22, impairing the function of endothelial cells (EC) and/or mural cells and leading to deficits in blood vessel development such as abnormal aortic arch morphology, tortuous retinal vessels, and tetralogy of Fallot. The mechanism by which dysfunctional endothelial cells and pericytes contribute to the vasculopathy, however, remains unknown. In this study, we used human blood vessel organoids (VOs) generated from iPSC of 22q11.2 DS patients to model the vascular malformations and genetic dysfunctions. We combined high-resolution lightsheet imaging and single-cell transcriptome analysis to link the genetic profile and vascular phenotype at the single-cell level. We developed a comprehensive analytical methodology by integrating deep learning-mediated blood vessel segmentation, network graph construction, and tessellation analysis for automated morphology characterization. We report that 22q11.2DS VOs demonstrate a smaller size with increased angiogenesis/sprouting, suggesting a less stable vascular network. Overall, clinical presentations of smaller vascular diameter, less connected vasculature, and increased branch points were recapitulated in 22q11.2DS VOs. Single-cell transcriptome profiling showed heterogeneity in both 22q11.2DS and control VOs, but the former demonstrated alterations in endothelial characteristics that are organ-specific and suggest a perturbation in the vascular developmental process. Intercellular communication analysis indicated that the vascular dysfunctions in 22q11.2 deletion were due to a lower cell-cell contact and upregulated extracellular matrix organization involving collagen and fibronectin. Voronoi diagram-based tessellation analysis also indicated that the colocalization of endothelial tubes and mural cells was different between control and 22q11.2 VOs, indicating that alterations in EC and mural interactions might contribute to the deficits in vascular network formation. This study illustrates the utility of VO in revealing the pathogenesis of 22q11.2DS vasculopathy.

bioengineering↗

bioRxiv2021

Transfer learning framework for cell segmentationwith incorporation of geometric features

With recent advances in multiplexed imaging and spatial transcriptomic and proteomic technologies, cell segmentation is becoming a crucial step in biomedical image analysis. In recent years, Fully Convolutional Networks (FCN) have achieved great success in nuclei segmentation in in vitro imaging. Nevertheless, it remains challenging to perform similar tasks on in situ tissue images with more cluttered cells of diverse shapes. To address this issue, we propose a novel transfer learning, cell segmentation framework incorporating shape-aware features in a deep learning model, with multi-level watershed and morphological post-processing steps. Our results show that incorporation of geometric features improves generalizability to segmenting cells in in situ tissue images, using solely in vitro images as training data.

systems biology↗

bioRxiv2019

Stag1 and Stag2 regulate cell fate decisions in hematopoiesis through non-redundant topological control

Transcriptional regulators, including the cohesin complex member STAG2, are recurrently mutated in cancer. The role of STAG2 in gene regulation, hematopoiesis, and tumor suppression remains unresolved. We show Stag2 deletion in hematopoietic stem/progenitor cells (HSPC) results in altered hematopoietic function, increased self-renewal, and impaired differentiation. ChIP-sequencing revealed that while Stag2 and Stag1 can bind the same loci, a component of Stag2 binding sites are unoccupied by Stag1 even in Stag2-deficient HSPCs. While concurrent loss of Stag2 and Stag1 abrogated hematopoiesis, Stag2 loss alone decreased chromatin accessibility and transcription of lineage-specification genes, including Ebf1 and Pax5, leading to blunted HSPC commitment to the B-cell lineage. Our data illustrate a role for Stag2 in transformation and transcriptional dysregulation distinct from its shared role with Stag1 in chromosomal segregation.\n\nOne Sentence SummaryStag1 rescues topologically associated domains in the absence of Stag2, but cannot restore chromatin architecture required for hematopoietic lineage commitment

cell biology↗

bioRxiv2017

Single-Cell Immune Map of Breast Carcinoma Reveals Diverse Phenotypic States Driven by the Tumor Microenvironment

Knowledge of immune cell phenotypes in the tumor microenvironment is essential for understanding mechanisms of cancer progression and immunotherapy response. We created an immune map of breast cancer using single-cell RNA-seq data from 45,000 immune cells from eight breast carcinomas, as well as matched normal breast tissue, blood, and lymph node. We developed a preprocessing pipeline, SEQC, and a Bayesian clustering and normalization method, Biscuit, to address computational challenges inherent to single-cell data. Despite significant similarity between normal and tumor tissue-resident immune cells, we observed continuous tumor-specific phenotypic expansions driven by environmental cues. Analysis of paired single-cell RNA and T cell receptor (TCR) sequencing data from 27,000 additional T cells revealed the combinatorial impact of TCR utilization on phenotypic diversity. Our results support a model of continuous activation in T cells and do not comport with the macrophage polarization model in cancer, with important implications for characterizing tumor-infiltrating immune cells.

immunology↗

bioRxiv2017

Heterogeneity of normal human breast stem and progenitor cells as revealed by transcriptional profiling

During development and pregnancy, the human mammary gland undergoes extensive remodeling in processes driven by populations of stem and progenitor cells. We recently reported that breast cancers are also hierarchically organized and driven by distinct populations of cancer stem cells characterized as CD44+CD24low/- or by expression of Aldehyde dehydrogenase (ALDH). These sets of markers identify largely non-overlapping mesenchymal and epithelial populations, each of which is capable of tumor initiation when transplanted into immunosuppressed mice. Less is known about these two populations, individually or their overlap, in the normal human mammary gland. The goal of this study was to understand the biology of the ALDH+ and CD44+CD24- populations in the normal human breast, using flow cytometry based sorting paired with functional ex vivo analyses, RNA-sequencing, and single cell RNA expression profiling. ALDH+ cells and ALDH-CD44+CD24- cells, generally, have epithelial-like and mesenchymal-like characteristics, respectively. Despite this, there are substantial similarities in the biological pathways activated in both populations when compared to differentiated cells. Additionally, we found a substantial proportion of cells that simultaneously express ALDH+ and CD44+CD24- whose abundance varies between individuals. At the single cell level, these cells have the greatest mammosphere forming capacity and express high levels of stemness and EMT-associated genes including ID1, SOX2, TWIST1, and ZEB2. Through unbiased analysis of individual ALDH+ cells, we find cells with either epithelial or mesenchymal expression phenotypes. We also identify a subpopulation of cells with a hybrid epithelial/mesenchymal expression phenotype that overexpress genes associated with aggressive triple negative breast cancers. These results highlight the utility of single cell analyses to characterize tissue heterogeneity, even in marker enriched cell populations, and further identifies the genes and pathways that define this heterogeneity.

developmental biology↗