Biology subjects

Das, J.

Publications and source records attributed to Das, J..

At least 73 records · Page 4Linked to original sources

bioRxiv2023

Stability and heterogeneity in the anti-microbiota reactivity of human milk-derived Immunoglobulin A

Immunoglobulin A (IgA) is secreted into breast milk and is critical to both protecting against enteric pathogens and shaping the infant intestinal microbiota. The efficacy of breast milk-derived maternal IgA (BrmIgA) is dependent upon its specificity, however heterogeneity in BrmIgA binding ability to the infant microbiota is not known. Using a flow cytometric array, we analyzed the reactivity of BrmIgA against bacteria common to the infant microbiota and discovered substantial heterogeneity between all donors, independent of preterm or term delivery. We also observed intra-donor variability in the BrmIgA response to closely related bacterial isolates. Conversely, longitudinal analysis showed that the anti-bacterial BrmIgA reactivity was relatively stable through time, even between sequential infants, indicating that mammary gland IgA responses are durable. Together, our study demonstrates that the anti-bacterial BrmIgA reactivity displays inter-individual heterogeneity but intra-individual stability. These findings have important implications for how breast milk shapes the development of the infant microbiota and protects against Necrotizing Enterocolitis. SummaryWe analyze the ability of breast milk-derived Immunoglobulin A (IgA) antibodies to bind the infant intestinal microbiota. We discover that each mother secretes into their breast milk a distinct set of IgA antibodies that are stably maintained over time.

immunology↗

bioRxiv2023

Population-level comparisons of gene regulatory networks modeled on high-throughput single-cell transcriptomics data.

Single-cell technologies enable high-resolution studies of phenotype-defining molecular mechanisms. However, data sparsity and cellular heterogeneity make modeling biological variability across single-cell samples difficult. We present SCORPION, a tool that uses a message-passing algorithm to reconstruct comparable gene regulatory networks from single cell/nuclei RNA-seq data that are suitable for population-level comparisons by leveraging the same baseline priors. Using synthetic data, we found that SCORPION outperforms 12 other gene regulatory network reconstruction techniques. Using supervised experiments, we show that SCORPION can accurately identify differences in regulatory networks between wild-type and transcription factor-perturbed cells. We demonstrate SCORPIONs scalability to population-level analyses using a single-cell RNA-seq atlas containing 200,436 cells from colorectal cancer and adjacent healthy tissues. The differences detected by SCORPION between tumor regions are consistent across population cohorts, as well as with our understanding of disease progression and elucidate phenotypic regulators that may impact patient survival.

bioinformatics↗

bioRxiv2023

A Multiscale Protein Abundance Structured Population Kinetic Model Systematically Explores the Design Space of Constitutive and Inducible CAR-T cells

Engineered chimeric antigen receptor (CAR)-T cells are designed to bind to antigens overexpressed on the surface of tumor cells and induce tumor cell lysis. However, healthy cells can express these antigens at lower abundances and can get lysed by CAR-T cells. A wide variety of CAR-T cells have been designed that increase tumor cell elimination while decreasing destruction of healthy cells. However, given the cost and labor-intensive nature of such efforts, a systematic exploration of potential hypotheses becomes limited. To this end, we develop a framework (PASCAR) by combining multiscale population dynamic models and multi-objective optimization approaches with data obtained from published cytometry and cytotoxicity assays to systematically explore design space of constitutive and tunable CAR-T cells. We demonstrate PASCAR can quantitatively describe in vitro and in vivo results for constitutive and inducible CAR-T cells and can successfully predict experiments outside the training data. Our exploration of the CAR design space reveals that CAR affinities in an intermediate range of dissociation constants (KD) in constitutive and tunable CAR-T cells can dramatically decrease healthy cell lysis but sustain a high rate of tumor cell killing. In addition, our modeling provides guidance towards optimal tuning of CAR expressions in synNotch CAR T cells. The proposed framework can be extended for other CAR immune cells.

systems biology↗

bioRxiv2022

BioNetGMMFit: a Parameter Estimation Tool for BioNetGen using Single-Cell Snapshot Data from Cell Populations Evolving over Time

BackgroundMechanistic models are commonly employed to describe signaling and gene regulatory kinetics in single cells and cell populations. Recent advances in single-cell technologies have produced multidimensional datasets where snapshots of copy numbers (or abundances) of a large number of proteins and mRNA are measured across time in single cells. The availability of such datasets presents an attractive scenario where mechanistic models are validated against experiments, and estimated model parameters enable quantitative predictions of signaling or gene regulatory kinetics. To empower the systems biology community to easily estimate parameters accurately from multidimensional single-cell data, we have merged a widely used rule-based modeling software package BioNetGen, which provides a user-friendly way to code for mechanistic models describing biochemical reactions, and the recently introduced CyGMM, that uses cell-to-cell differences to improve parameter estimation for such networks, into a single software package: BioNetGMMFit. ResultsBioNetGMMFit provides parameter estimates of the model, supplied by the user in the BioNetGen markup language (BNGL), which yield the best fit for the observed single-cell, timestamped data of cellular components. Furthermore, for more precise estimates, our software generates confidence intervals around each model parameter. BioNetG-MMFit is capable of fitting datasets of increasing cell population sizes for any mechanistic model specified in the BioNetGen markup language. ConclusionBy streamlining the process of developing mechanistic models for large single-cell datasets, BioNetGMMFit provides an easily-accessible modeling framework designed for scale and the broader biochemical signaling community.

systems biology↗

bioRxiv2022

SLIDE: Significant Latent Factor Interaction Discovery and Exploration across biological domains

Modern multi-omic technologies can generate deep multi-scale profiles. However, differences in data modalities, multicollinearity of the data, and large numbers of irrelevant features make the analyses and integration of high-dimensional omic datasets challenging. Here, we present Significant Latent factor Interaction Discovery and Exploration (SLIDE), a first-in-class interpretable machine learning technique for identifying significant interacting latent factors underlying outcomes of interest from high-dimensional omic datasets. SLIDE makes no assumptions regarding data-generating mechanisms, comes with theoretical guarantees regarding identifiability of the latent factors/corresponding inference, outperforms/performs at least as well as state-of-the-art approaches in terms of prediction, and provides inference beyond prediction. Using SLIDE on scRNA-seq data from systemic sclerosis (SSc) patients, we first uncovered significant interacting latent factors underlying SSc pathogenesis. In addition to accurately predicting SSc severity and outperforming existing benchmarks, SLIDE uncovered significant factors that included well-elucidated altered transcriptomic states in myeloid cells and fibroblasts, an intriguing keratinocyte-centric signature validated by protein staining, and a novel mechanism involving altered HLA signaling in myeloid cells, that has support in genetic data. SLIDE also worked well on spatial transcriptomic data and was able to accurately identify significant interacting latent factors underlying immune cell partitioning by 3D location within lymph nodes. Finally, SLIDE leveraged paired scRNA-seq and TCR-seq data to elucidate latent factors underlying extents of clonal expansion of CD4 T cells in a nonobese diabetic model of T1D. The latent factors uncovered by SLIDE included well-known activation markers, inhibitory receptors and intracellular regulators of receptor signaling, but also honed in on several novel naive and memory states that standard analyses missed. Overall, SLIDE is a versatile engine for biological discovery from modern multi-omic datasets.

systems biology↗

bioRxiv2022

A gut commensal protist protects against virus-mediated loss of oral tolerance

Loss of oral tolerance (LOT) to gluten, characterized by a T helper 1 (Th1) gluten-specific immune response, is a hallmark of celiac disease (CeD) and can be triggered by enteric viral infections. We hypothesized that certain gut microbes have the capacity to protect against virus-mediated LOT. By using our previously defined reovirus-mediated LOT CeD model, we discovered that the gut colonizing protist Tritrichomonas (T.) arnold promotes oral tolerance and protects against reovirus-mediated LOT by suppressing the reovirus-induced proinflammatory program of dietary-antigen-presenting CD103+ dendritic cells. Importantly, T. arnold did not affect antiviral host immunity, suggesting that T. arnold-mediated protection against T1L-induced LOT is not attributable to differences in antiviral host responses. Additionally, using gnotobiotic mice, we found that Tritrichomonas arnold colonization is sufficient to protect against reovirus-mediated LOT in the absence of the microbiota. Mechanistically, we show that Tritrichomonas arnold colonization restrains reovirus-induced inflammatory responses in dendritic cells and thus limit their ability to promote Th1 immune responses ex vivo. Finally, our studies using human stool samples support a role for Tritrichomonas sp. colonization in protecting against development of CeD. This study will motivate the design of effective therapies to prevent LOT to gluten in at-risk individuals and to reinstate tolerance to gluten in CeD patients. One Sentence SummaryTritrichomonas arnold protects against virus-mediated loss of oral tolerance to gluten and is underrepresented in celiac disease patients.

immunology↗

bioRxiv2022

Probing the role of macromolecular crowding in cell volume regulation using fluorescence anisotropy

Cytoplasmic macromolecular crowding (MMC) influences multiple cellular functions. Using fluorescence anisotropy of EGFP, we explore the homeostasis of MMC in the cytoplasm and its role in cell volume regulation. Individual cells of different lineages maintain a distinctly unique average cytoplasmic MMC for considerably long durations. Despite diffusion, actin cytoskeleton facilitates spatially heterogeneous MMC in the cytoplasm. While hypertonic-stress triggers regulatory volume increase (RVI), other methods of increasing cytoplasmic MMC fail to start RVI, suggesting that cells lack capabilities for MMC homeostasis. During spreading or microtubule-depolymerized state, cells neglect changes in cytoplasmic MMC to undergo volume regulation. Inhibition of TNFR1 increases the membrane tension and deprives cells of their ability to undergo volume regulation. Current understanding contemplates that cytoplasmic MMC, the mechanical state of the plasma membrane or the actin cytoskeleton, could be sensors for cell volume regulation. Our observations establish the irrelevance of cytoplasmic MMC and the significance of plasma membrane tension in setting off the cell volume regulation machinery.

cell biology↗

bioRxiv2022

Generalized Method of Moments improves parameter estimation in biochemical signaling models of time-stamped single-cell snapshot data

MotivationOrdinary differential equations are commonly used to model the sub-cellular dynamics of average values of proteins and mRNAs. New single-cell technologies provide cell-to-cell differences in protein/mRNA abundances that allow for the evaluation of higher order moments. However, using this additional information to improve parameter estimation can be challenging since the magnitudes of single-cell abundances can vary widely between proteins/mRNA. ResultsWe employ Generalized Method of Moments (GMM) and Particle Swarm Optimization to address the above challenges in mechanistic modeling of signaling kinetics data. Using synthetic data from linear and non-linear models, we show that the proposed method improves parameter estimation and enables construction of approximate confidence intervals. Furthermore, our approach exploits parallel computation to scale with increasing data size and dimensions. We apply our software CyGMM to estimate parameters in a linear ODE model for publicly available longitudinal single-cell cytometry data for CD8+ T cells. Our results demonstrate substantial improvements for modeling data from single-cell cytometry and RNA-seq experiments. AvailabilityWe also make freely available our estimation software CyGMM written in C++ on github (https://github.com/jhnwu3/CyGMM). Contactjayajit@gmail.com Supplementary informationSupplementary data are available at Bioinformatics online.

systems biology↗

bioRxiv2022

The Olfm4-defined human neutrophil subsets differ in proteomic profile in septic shock

The specific granule glycoprotein olfactomedin-4 (Olfm4) marks a constitutive subset of neutrophils in humans, where 1-70% of peripheral blood neutrophils produce Olfm4. The proportion of Olfm4-high (Olfm4-H) neutrophils correlates with the severity of paediatric septic shock and could predict mortality in adult septic shock in previous studies, but it is not known whether and how the Olfm4-H neutrophils contribute to sepsis pathogenesis. The aim of this study was to decipher proteomic differences between the Olfm4-H and Olfm4-low (Olfm4-L) human neutrophil subsets at baseline and in the context of septic shock, hypothesizing that Olfm4 marks a neutrophil subset with a distinct proteomic profile, predisposing it for detrimental processes in sepsis. A novel protocol for the preparation of fixed, antibody-stained and sorted neutrophils for LC-MS/MS analysis of proteome was developed. In neutrophil subsets from healthy blood donors, 47 proteins had significantly higher abundance in the Olfm4-H population, and 62 proteins in the Olfm4-L population. Pathway enrichment analysis showed that the differences concerned proteins related to neutrophil degranulation, with e.g. Rab3d and a subunit of the vacuolar ATPase proton pump being more abundant in the Olfm4-H neutrophils, and the alarmin S100-A7, the major neutrophil chemotactic receptor CXCR1 and the antimicrobial peptide defensin alpha-4 being more abundant in the Olfm4-L neutrophils. The data suggest different preparedness to infection in the subsets. In the limited material analysed here, there was no significant correlation between the severity of sepsis and the proportion of Olfm4-H neutrophils, but an increased concentration of Olfm4 in plasma from septic shock patients as compared to healthy blood donors was observed. Furthermore, in neutrophil subsets isolated from septic shock patients, 28 proteins had significantly higher abundance in the Olfm4-H subset and 38 in the Olfm4-L subset, the latter including e.g. Fc receptor proteins and MHC class I molecules, suggesting distinct immunological responses. This is the first report pointing towards differential functions of the Olfm4-defined neutrophil subpopulations in humans and the data are consistent with the idea of distinct responses in the subsets during infection and inflammation.

immunology↗

bioRxiv2022

Patterns of Volatility Across the Spike Protein Accurately Predict the Emergence of Mutations within SARS-CoV-2 Lineages

Mutations in the spike glycoprotein of SARS-CoV-2 allow the virus to probe the sequence space in search of higher-fitness states. New sublineages of SARS-CoV-2 variants-of-concern (VOCs) continuously emerge with such mutations. Interestingly, the sites of mutation in these sublineages vary between the VOCs. Whether such differences reflect the random nature of mutation appearance or distinct evolutionary spaces of spike in the VOCs is unclear. Here we show that each position of spike has a lineage-specific likelihood for mutations to appear and dominate descendent sublineages. This likelihood can be accurately estimated from the lineage-specific mutational profile of spike at a protein-wide level. The mutability environment of each position, including adjacent sites on the protein structure and neighboring sites on the network of comutability, accurately forecast changes in descendent sublineages. Mapping of imminent changes within the VOCs can contribute to the design of immunogens and therapeutics that address future forms of SARS-CoV-2.

bioinformatics↗

bioRxiv2022

Antibodies targeting conserved non-canonical antigens and endemic coronaviruses associate with favorable outcomes in severe COVID-19

While there have been extensive analyses characterizing cellular and humoral responses across the severity spectrum in COVID-19, predictors of outcomes within severe COVID-19 remain to be comprehensively elucidated. Recently, we identified divergent monocyte states as predictors of outcomes within severe COVID-19, but corresponding humoral profiles of risk have not been delineated. Furthermore, the nature of antibodies (Abs) directed against viral antigens beyond the spike protein or endemic coronavirus antigens and their associations with disease severity and outcomes remain poorly defined. We performed deep molecular profiling of Abs directed against a wide range of antigenic specificities in severe COVID-19 patients admitted to the ICU. The profiles consisted of canonical (S, RBD, N) and non-canonical (orf3a, orf8, nsp3, nps13 and M) antigenic specificities. Notably, multivariate machine learning (ML) models, generated using profiles of Abs directed against canonical or non-canonical antigens, were equally discriminative of recovery and mortality COVID-19 outcomes. In both ML models, survivors were associated with increased virus-specific IgA and IgG3 antibodies and with higher antigen-specific antibody galactosylation. Intriguingly, pre-pandemic healthy controls had cross-reactive Abs directed against nsp13 which is a conserved protein in other alpha and beta coronaviruses. Notably, higher levels of nsp13-specific IgA antibodies were associated with recovery in severe COVID-19. In keeping with these findings, a model built on Ab profiles for endemic coronavirus antigens was also predictive of COVID-19 outcome bifurcation, with higher levels of IgA and IgG3 antibodies against OC43 S and NL63 S being associated with survival. Our results suggest the importance of Abs targeting non-canonical SARS-CoV-2 antigens as well as those directed against endemic coronaviruses in favorable outcomes of severe COVID-19.

immunology↗

bioRxiv2021

McSNAC: A software to approximate first-order signaling networks from mass cytometry data

Mass cytometry (CyTOF) gives unprecedented opportunity to simultaneously measure up to 40 proteins in single cells, with a theoretical potential to reach 100 proteins. This high-dimensional single-cell information can be very useful to dissecting mechanisms of cellular activity. In particular, measuring abundances of signaling proteins like phospho-proteins can provide detailed information on the dynamics of single-cell signaling processes. With a proper computational analysis, timestamped CyTOF data of signaling proteins could help develop predictive and mechanistic models for signaling kinetics. These models would be useful for predicting the effects of perturbations in cells, or comparing signaling networks across cell groups. We propose our Mass cytometry Signaling Network Analysis Code, or McSNAC, a new software capable of reconstructing signaling networks and estimating their kinetic parameters from CyTOF data. McSNAC approximates signaling networks as a network of first-order reactions between proteins. This assumption breaks down often as signaling reactions can involve binding and unbinding, enzymatic reactions, and other nonlinear constructions. Furthermore, McSNAC may be limited to approximating indirect interactions between protein species, as cytometry experiments are only able to assay a small fraction of the protein species that are involved in signaling. We carry out a series of in silico experiments here to show that 1) McSNAC is capable of accurately estimating the ground-truth model in a scalable manner when given data originating from a first-order system; 2) McSNAC is capable of qualitatively predicting outcomes to perturbations of species abundances in simple second-order reaction models and in a complex in silico nonlinear signaling network in which some proteins are unmeasured. These findings demonstrate that McSNAC can be a valuable screening tool for generating models of signaling networks from timestamped CyTOF data.

systems biology↗

bioRxiv2021

Spatially Resolved In Silico Modeling of NKG2D Signaling Kinetics Reveals Key role of NKG2D and Vav1 Co-clustering in Generating Natural Killer cell Activation

Natural Killer (NK) cells provide key resistance against viral infections and tumors. A diverse set of activating and inhibitory NK cell receptors (NKRs) interact with cognate ligands presented by target host cells, where integration of dueling signals initiated by the ligand-NKR interactions determines NK cell activation or tolerance. Imaging experiments over decades have shown micron and sub-micron scale spatial clustering of activating and inhibitory NKRs. The mechanistic roles of these clusters in affecting downstream signaling and activation are often unclear. To this end, we developed a predictive in silico framework by combining spatially resolved mechanistic agent based modeling, published TIRF imaging data, and parameter estimation to determine mechanisms by which formation and spatial movements of activating NKG2D microclusters affect early time NKG2D signaling kinetics in a human cell line NKL. We show co-clustering of NKG2D and the guanosine nucleotide exchange factor Vav1 in NKG2D microclusters plays a dominant role over ligand (ULBP3) rebinding in increasing production of phospho-Vav1(pVav1), an activation marker of early NKG2D signaling. The in silico model successfully predicts several scenarios of inhibition of NKG2D signaling and time course of NKG2D spatial clustering over a short ([~]3 min) interval. Modeling shows the presence of a spatial positive feedback relating formation and centripetal movements of NKG2D microclusters, and pVav1 production offers flexibility towards suppression of activating signals by inhibitory KIR ligands organized in inhomogeneous spatial patterns (e.g., a ring). Our in silico framework marks a major improvement in developing spatiotemporal signaling models with quantitatively estimated model parameters using imaging data.

systems biology↗

bioRxiv2021

CD4+CCR6+ T cells dominate the BCG-induced transcriptional signature

The century-old Mycobacterium bovis Bacillus Calmette-Guerin (BCG) remains the only licensed vaccine against tuberculosis (TB). Despite this, there is still a lot to learn about the immune response induced by BCG, both in terms of phenotype and specificity. Here, we investigated immune responses in adult individuals pre and 8 months post BCG vaccination. We specifically determined changes in gene expression, cell subset composition, DNA methylome, and the TCR repertoire induced in PBMCs and CD4 memory T cells associated with antigen stimulation by either BCG or a Mycobacterium tuberculosis (Mtb)-derived peptide pool. Following BCG vaccination, we observed increased frequencies of CCR6+ CD4 T cells, which includes both Th1* and Th17 subsets, and mucosal associated invariant T cells (MAITs). A large number of immune response genes and pathways were upregulated post BCG vaccination with similar patterns observed in both PBMCs and memory CD4 T cells, thus suggesting a substantial role for CD4 T cells in the cellular response to BCG. These upregulated genes and associated pathways were also reflected in the DNA methylome. We described both qualitative and quantitative changes in the BCG-specific TCR repertoire post vaccination, and importantly found evidence for similar TCR repertoires across different subjects. The immune signatures defined herein can be used to track and further characterize immune responses induced by BCG, and can serve as reference for benchmarking novel vaccination strategies.

immunology↗

bioRxiv2021

Network-based integration of epigenetic landscapes unveils molecular programs underlying human T follicular helper cell differentiation

T follicular helper (Tfh) cells play a critical role in T-dependent humoral immune responses. While genetic programs controlling Tfh cell differentiation have been extensively studied using murine models, studies in humans have been hampered by the lack of a robust in vitro differentiation system for Tfh cells. We characterized epigenomic landscapes across stages of Tfh cell differentiation in a healthy human tonsil using ATAC-Seq and CUT&RUN for selected histone modifications. We combined these epigenomic datasets and integrated them with the reference human protein interactome using a novel network propagation approach. Our approach uncovered subnetworks integral to Tfh cell differentiation. These subnetworks captured known Tfh cell drivers to a greater extent than conventional gene-centric analyses would, and also revealed novel modules that may be required for Tfh cell differentiation. We find that human Tfh cell subnetworks are functionally associated with specific immune signaling cascades including cytokine receptor driven pathways. Analyses of transcriptomic data revealed that in addition to these immune pathways being significantly dysregulated in severe COVID-19, the corresponding Tfh cell subnetworks are also transcriptionally perturbed to a similar extent. This provides a molecular mechanistic basis for the previously observed impaired Tfh cell differentiation and loss of germinal centers in severe COVID-19.

immunology↗

bioRxiv2021

Expression of GRINA Correlates with Prognosis in Human Cancers: A Pan-cancer Analysis

GRINA is an emerging target for cancer therapy. However, the role of GRINA expression and its correlation with cancer patient survival has not been comprehensively studied. Here, we found that mRNA and protein expression of GRINA was upregulated in breast, colon, gastric, and prostate cancers and negatively correlated with patient survival. Also, the upregulation of GRINA expression is associated with hypomethylation of its promoter region. Our GRINA-miRNAs network analysis revealed potential regulatory miRNAs regulating the GRINA expression and its downstream pathways. Next, functional enrichment and pathway analysis of genes commonly co-express with GRINA in breast, colon, gastric, and prostate cancers revealed GRINA regulatory pathways. Concurrently, our upstream regulator analysis revealed possible kinases, transcription factors, and proteins that may potentially regulate GRINA. Overall, this study demonstrates the prognostic significance of GRINA expression and identifies potential regulatory mechanisms, which might have significant implications in targeted therapies for human cancers.

cancer biology↗

bioRxiv2021

CausER - a framework for inferring causal latent factors using multi-omic human datasets

High-dimensional cellular and molecular profiling of human samples highlights the need for analytical approaches that can integrate multi-omic datasets to generate predictive biomarkers and prioritized causal inferences. Current methods are limited by high dimensionality of the combined datasets, the differences in their data distributions and their integration to infer causal relationships. Here we present Essential Regression (ER), an interpretable machine learning approach for high-dimensional multi-omic datasets, that addresses these problems by identifying latent factors and their likely cause-effect relationships with the system-wide outcome/properties of interest. ER is a novel data-distribution-free latent-factor regression model that integrates multi-omic datasets and identifies latent factors significantly associated with an outcome. ER outperforms a range of state-of-the-art methods in terms of prediction performance on simulated datasets. ER can be coupled with probabilistic graphical modeling thereby strengthening the causal inferences. ER generates novel cellular and molecular predictions, using multi-omic human systems immunology datasets, pertaining to immunosenescence and immune dysregulation.

systems biology↗

bioRxiv2021

Determining clinically relevant features in cytometry data using persistent homology

Cytometry experiments yield high-dimensional point cloud data that is difficult to interpret manually. Boolean gating techniques coupled with comparisons of relative abundances of cellular subsets is the current standard for cytometry data analysis. However, this approach is unable to capture more subtle topological features hidden in data, especially if those features are further masked by data transforms or significant batch effects or donor-to-donor variations in clinical data. We present that persistent homology, a mathematical structure that summarizes the topological features, can distinguish different sources of data, such as from groups of healthy donors or patients, effectively. Analysis of publicly available cytometry data describing non-naive CD8+ T cells in COVID-19 patients and healthy controls shows that systematic structural differences exist between single cell protein expressions in COVID-19 patients and healthy controls. Our method identifies proteins of interest by a decision-tree based classifier and passes them to a kernel-density estimator (KDE) for sampling points from the density distribution. We then compute persistence diagrams from these sampled points. The resulting persistence diagrams identify regions in cytometry datasets of varying density and identify protruded structures such as elbows. We compute Wasserstein distances between these persistence diagrams for random pairs of healthy controls and COVID-19 patients and find that systematic structural differences exist between COVID-19 patients and healthy controls in the expression data for T-bet, Eomes, and Ki-67. Further analysis shows that expression of T-bet and Eomes are significantly downregulated in COVID-19 patient non-naive CD8+ T cells compared to healthy controls. This counter-intuitive finding may indicate that canonical effector CD8+ T cells are less prevalent in COVID-19 patients than healthy controls. This method is applicable to any cytometry dataset for discovering novel insights through topological data analysis which may be difficult to ascertain otherwise with a standard gating strategy or in the presence of large batch effects. Author summaryIdentifying differences between cytometry data seen as a point cloud can be complicated by random variations in data collection and data sources. We apply persistent homology used in topological data analysis to describe the shape and structure of the data representing immune cells in healthy donors and COVID-19 patients. By looking at how the shape and structure differ between healthy donors and COVID-19 patients, we are able to definitively conclude how these groups differ despite random variations in the data. Furthermore, these results are novel in their ability to capture shape and structure of cytometry data, something not described by other analyses.

systems biology↗