High-dimensional Single-cell Data Research Articles

NK Cell Diversity Influences Outcomes after Cord Blood Transplantation

Natural killer (NK) cells are the first subset of lymphocytes to reconstitute, and likely play an important role in the graft-versus-leukemia response early after allogeneic hematopoietic stem cell transplantation. NK cells are highly diverse, and express an array of activating and inhibiting receptors, which delicately regulate their activation status.To provide a framework to better understand NK cell development and maturation after cord blood transplantation (CBT) and its implication on clinical outcomes, we designed a mass cytometry (also known as cytometry by time of flight) panel of 40 monoclonal antibodies recognizing lineage markers, NK cell receptors and transcription factors to interrogate the NK cell repertoire in prospectively collected peripheral blood (PB) samples from 60 CBT recipients. The resultant high-dimensional single cell data were analyzed by t-Distributed Stochastic Neighbor Embedding (t-SNE) dimension reduction algorithm to generate expression profile graphs.With this mass cytometry platform, we first interrogated and compared the NK cell repertoire in CB (n=10) and adult peripheral blood (PB) (n=20) and used inverse Simpson index to measure the diversity of the NK cell populations. NK cells from CB donors had a significantly lower diversity (median 537, range 311-891) compared to NK cells from healthy adult PB donors (median 2858, range 1228-5630; p<0.0001). Since CMV infection has been shown to drive NK cell maturation, we next asked if CMV infection also alters NK cell diversity by comparing the diversity of the NK cell repertoire in CB (n=10), and in PB from adult healthy donors who were either CMV-seropositive (n=10), or CMV-seronegative (n=10). The diversity of the NK cell repertoire was significantly greater in CMV seropositive healthy individuals (median 4217, range 2672-5630) compared to CMV-seronegative healthy adults (median 2247 , range 2026-2468; P=0.0002) and cord blood donors (median 537, range 311-891; P<0.0001), pointing to an important role for CMV infection in driving NK diversity.Harnessing mass cytometry, we undertook an in-depth examination of the NK repertoire in prospectively collected PB samples from 60 CBT recipients. We first sought to determine if CMV infection early after CBT can drive NK cell maturation and accelerate the acquisition of a diverse repertoire. The median time to CMV reactivation for the cohort was 71 days. CMV reactivation in the first 100 days after CBT (n=25) was associated with significantly greater NK diversity (median 844, range 198-1196) when compared to patients without CMV reactivation (median 405, range 110-889; p=0.0003). Notably, CMV augmented NK diversity mainly in the KIR+ population (median 743, range 207-1500) compared to the KIR- population (median 361, range 92-1038; p=0.007); while in the absence of CMV reactivation, NK cell diversity in the KIR+ and KIR- populations remained similar (median 305, range 102-702 vs. median 332, range 54-830; p=0.98). These data indicate that CMV infection augments NK diversity after CBT, and that this effect is most pronounced within the KIR+ population.We next determined if NK cell diversity can influence outcomes by comparing the diversity of the NK cell repertoire in PB samples collected in the first 100 days after CBT from 15 patients who went on to relapse compared to 45 patients who remained in remission for 2 years. NK cell diversity was significantly higher (median 839, range 416-1069) in patients who remained in remission compared to those who progressed and relapsed (median 446, range 328-925; p=0.03). Moreover, the greatest difference in diversity in the two groups was noted within the KIR+ population (median 788, range 101-1500 for patients in remission vs median 320, range 177-857 for patients who relapsed; p=0.002).In summary, we have used multiparametric mass cytometry to examine expression of 40 NK markers, and to gain insights into the diversity of the NK cell repertoire after CBT. We have shown that CB NK cells have a limited NK cell diversity, which increases in breadth and subsets after transplant. Moreover, we observed that CMV infection reconfigures the NK cell compartment and drives the diversity of the NK cell repertoire, especially within the KIR+ population. These data may provide the basis to rationally design therapeutic strategies that can modulate NK diversity in the settings of infection and transplant. DisclosuresNo relevant conflicts of interest to declare.

CytoPy: An autonomous cytometry analysis framework.

Cytometry analysis has seen a considerable expansion in recent years in the maximum number of parameters that can be acquired in a single experiment. In response to this technological advance there has been an increased effort to develop new computational methodologies for handling high-dimensional single cell data acquired by flow or mass cytometry. Despite the success of numerous algorithms and published packages to replicate and outperform traditional manual analysis, widespread adoption of these techniques has yet to be realised in the field of immunology. Here we present CytoPy, a Python framework for automated analysis of cytometry data that integrates a document-based database for a data-centric and iterative analytical environment. In addition, our algorithm-agnostic design provides a platform for open-source cytometry bioinformatics in the Python ecosystem. We demonstrate the ability of CytoPy to phenotype T cell subsets in whole blood samples even in the presence of significant batch effects due to technical and user variation. The complete analytical pipeline was then used to immunophenotype the local inflammatory infiltrate in individuals with and without acute bacterial infection. CytoPy is open-source and licensed under the MIT license. CytoPy is available at https://github.com/burtonrj/CytoPy, with notebooks accompanying this manuscript (https://github.com/burtonrj/CytoPyManuscript) and software documentation at https://cytopy.readthedocs.io/.

Machine Learning Approaches Identify Genes Containing Spatial Information From Single-Cell Transcriptomics Data.

The development of single-cell sequencing technologies has allowed researchers to gain important new knowledge about the expression profile of genes in thousands of individual cells of a model organism or tissue. A common disadvantage of this technology is the loss of the three-dimensional (3-D) structure of the cells. Consequently, the Dialogue on Reverse Engineering Assessment and Methods (DREAM) organized the Single-Cell Transcriptomics Challenge, in which we participated, with the aim to address the following two problems: (a) to identify the top 60, 40, and 20 genes of the Drosophila melanogaster embryo that contain the most spatial information and (b) to reconstruct the 3-D arrangement of the embryo using information from those genes. We developed two independent techniques, leveraging machine learning models from least absolute shrinkage and selection operator (Lasso) and deep neural networks (NNs), which are applied to high-dimensional single-cell sequencing data in order to accurately identify genes that contain spatial information. Our first technique, Lasso.TopX, utilizes the Lasso and ranking statistics and allows a user to define a specific number of features they are interested in. The NN approach utilizes weak supervision for linear regression to accommodate for uncertain or probabilistic training labels. We show, individually for both techniques, that we are able to identify important, stable, and a user-defined number of genes containing the most spatial information. The results from both techniques achieve high performance when reconstructing spatial information in D. melanogaster and also generalize to zebrafish (Danio rerio). Furthermore, we identified novel D. melanogaster genes that carry important positional information and were not previously suspected. We also show how the indirect use of the full datasets’ information can lead to data leakage and generate bias in overestimating the model’s performance. Lastly, we discuss the applicability of our approaches to other feature selection problems outside the realm of single-cell sequencing and the importance of being able to handle probabilistic training labels. Our source code and detailed documentation are available at https://github.com/TJU-CMC-Org/SingleCell-DREAM/.

Mass Cytometry Analysis of T-Helper Cells.

CD4+ T cells or helper T cells play various roles in the immune response to pathogens, tumors, as well as in asthma, allergy, and autoimmunity. Consequently, there is great interest in the comprehensive investigation of different T helper cell subsets. Here, we use mass cytometry (CyTOF), which is similar to flow cytometry but uses metal ion-tagged antibodies, which are detected using time-of-flight mass spectrometry. CyTOF allows the simultaneous detection of over 40 different antibodies, allowing us to collect high-dimensional single-cell proteomic data on T helper subsets. We use an extensive staining panel with a large number of lineage markers, cytokines, and other functional markers to identify and characterize CD4+ T cell subsets. In this method, human peripheral blood mononuclear cells are stimulated ex vivo with PMA and ionomycin, which activates T cells. Theactivated CD4+ T cells can then be identified as Th1, Th2, or Th17 cells based on their production of IFNγ, IL-4, and IL-17, respectively. Tregs are identified as CD4+CD25+CD127lo. Once Th1, Th2, Th17, and Tregs have been identified, they can be characterized in more detail using the large number of phenotypic and functional markers included in the CyTOF staining panel. Finally, automated and unbiased high-dimensional data analysis tools can be employed to comprehensively characterize T helper cells and discover novel features.

SCHNEL: scalable clustering of high dimensional single-cell data.

Single cell data measures multiple cellular markers at the single-cell level for thousands to millions of cells. Identification of distinct cell populations is a key step for further biological understanding, usually performed by clustering this data. Dimensionality reduction based clustering tools are either not scalable to large datasets containing millions of cells, or not fully automated requiring an initial manual estimation of the number of clusters. Graph clustering tools provide automated and reliable clustering for single cell data, but suffer heavily from scalability to large datasets. We developed SCHNEL, a scalable, reliable and automated clustering tool for high-dimensional single-cell data. SCHNEL transforms large high-dimensional data to a hierarchy of datasets containing subsets of data points following the original data manifold. The novel approach of SCHNEL combines this hierarchical representation of the data with graph clustering, making graph clustering scalable to millions of cells. Using seven different cytometry datasets, SCHNEL outperformed three popular clustering tools for cytometry data, and was able to produce meaningful clustering results for datasets of 3.5 and 17.2 million cells within workable time frames. In addition, we show that SCHNEL is a general clustering tool by applying it to single-cell RNA sequencing data, as well as a popular machine learning benchmark dataset MNIST. Implementation is available on GitHub (https://github.com/biovault/SCHNELpy). All datasets used in this study are publicly available. Supplementary data are available at Bioinformatics online.

A Single-Cell Transcriptional Roadmap of the Mouse and Human Lymph Node Lymphatic Vasculature.

Single-cell transcriptomics promise to revolutionize our understanding of the vasculature. Emerging computational methods applied to high-dimensional single-cell data allow integration of results between samples and species and illuminate the diversity and underlying developmental and architectural organization of cell populations. Here, we illustrate these methods in the analysis of mouse lymph node (LN) lymphatic endothelial cells (LEC) at single-cell resolution. Clustering identifies five well-delineated subsets, including two medullary sinus subsets not previously recognized as distinct. Nearest neighbor alignments in trajectory space position the major subsets in a sequence that recapitulates the known features and suggests novel features of LN lymphatic organization, providing a transcriptional map of the lymphatic endothelial niches and of the transitions between them. Differences in gene expression reveal specialized programs for (1) subcapsular ceiling endothelial interactions with the capsule connective tissue and cells; (2) subcapsular floor regulation of lymph borne cell entry into the LN parenchyma and antigen presentation; and (3) pathogen interactions and (4) LN remodeling in distinct medullary subsets. LEC of the subcapsular sinus floor and medulla, which represent major sites of cell entry and exit from the LN parenchyma respectively, respond robustly to oxazolone inflammation challenge with enriched signaling pathways that converge on both innate and adaptive immune responses. Integration of mouse and human single-cell profiles reveals a conserved cross-species pattern of lymphatic vascular niches and gene expression, as well as specialized human subsets and genes unique to each species. The examples provided demonstrate the power of single-cell analysis in elucidating endothelial cell heterogeneity, vascular organization, and endothelial cell responses. We discuss the findings from the perspective of LEC functions in relation to niche formations in the unique stromal and highly immunological environment of the LN.

PARC: ultrafast and accurate clustering of phenotypic data of millions of single cells.

MotivationNew single-cell technologies continue to fuel the explosive growth in the scale of heterogeneous single-cell data. However, existing computational methods are inadequately scalable to large datasets and therefore cannot uncover the complex cellular heterogeneity.ResultsWe introduce a highly scalable graph-based clustering algorithm PARC—Phenotyping by Accelerated Refined Community-partitioning—for large-scale, high-dimensional single-cell data (>1 million cells). Using large single-cell flow and mass cytometry, RNA-seq and imaging-based biophysical data, we demonstrate that PARC consistently outperforms state-of-the-art clustering algorithms without subsampling of cells, including Phenograph, FlowSOM and Flock, in terms of both speed and ability to robustly detect rare cell populations. For example, PARC can cluster a single-cell dataset of 1.1 million cells within 13 min, compared with >2 h for the next fastest graph-clustering algorithm. Our work presents a scalable algorithm to cope with increasingly large-scale single-cell analysis.Availability and implementation https://github.com/ShobiStassen/PARC.Supplementary information Supplementary data are available at Bioinformatics online.

Meeting the Challenges of High-Dimensional Single-Cell Data Analysis in Immunology.

Recent advances in cytometry have radically altered the fate of single-cell proteomics by allowing a more accurate understanding of complex biological systems. Mass cytometry (CyTOF) provides simultaneous single-cell measurements that are crucial to understand cellular heterogeneity and identify novel cellular subsets. High-dimensional CyTOF data were traditionally analyzed by gating on bivariate dot plots, which are not only laborious given the quadratic increase of complexity with dimension but are also biased through manual gating. This review aims to discuss the impact of new analysis techniques for in-depths insights into the dynamics of immune regulation obtained from static snapshot data and to provide tools to immunologists to address the high dimensionality of their single-cell data.

Abstract LB-218: High-dimensional modeling of single-cell-based CD8+ T cell exhaustion predicts response to immune checkpoint blockade

Abstract INTRODUCTION Accurate prediction of response to existing immune checkpoint blockade (ICB) and development of more efficacious ICB strategies are needed for cancer immunotherapy. The single-cell-level characterization of tumor-infiltrating CD8+ T lymphocytes (TILs) has presented new opportunities. However, the phenotypic variability of TILs, including tumor-specificity and CD8+ TIL exhaustion that regulate response to ICB, is poorly understood. Computationally modeling a large number of CD8+ TILs with high-dimensional T cell functional markers can link high-dimensional single-cell data with fundemental T cell functionalities. METHODS Here we develop a computational method (called HD-scMed) to predict response to ICB at the single-cell level. Our high-dimensional single-cell method, HD-scMed, models the phenotypic variability of CD8+ TILs directly within the original high-dimensional expression space, defined by k-dimensional exhaustion markers and other T cell functional markers. Within the high-dimensional expression space, HD-scMed identifies Pareto-optimal Frontier (PF) TILs representing a subset of exhausted TILs, by a mathematical model, Pareto Optimization. A quantitative criterion, D-value, defined as the Euclidean distance from PF TILs to the baseline TILs excluded by the PF, is used to indicate the extent to which the PF TILs are exhausted. RESULTS D-value accurately predicts clinical response to the ICB, i.e., αPD-1, αCTLA-4, and αPD-1+αCTLA-4, with performance of AUC=100% by single-cell flow cytometry (CyTOF) data with 101,401 CD8+ TILs from human tumors after ICBs, and AUC=95% and 86% by single-cell RNA-seq (scRNA-seq) data including 9,405 CD8+ TILs from human tumors before and after ICBs, respectively. Mechanistically, “D-high” TILs are associated with high tumor-specificity, enhanced exhaustion, and inactivated effector and cytotoxic signatures in non-responders; whereas “D-low” TILs are enriched in responders demonstrating tumor-specific T cells with T cell activation and cytolytic effector T cell signatures. We also observed “D negative” bystander T cells irrelevant to response. Notably, LAG3 shows significantly higher contribution to TIL D values in non-responders than responders. To reverse the functionality of the LAG3-high TILs after receiving αPD-1+αCTLA-4, we tested the effects of αLAG3 with αPD-1+αCTLA-4 in murine tumor models. Remarkably, the combination, αLAG3+αPD-1+αCTLA-4, showed extraordinary antitumor efficacy in eradicating advanced tumors, which was associated with burst GZMBhi TIL populations; whereas αPD-1+αCTLA-4 or αPD-1+αCTLA-4 plus other ICBs as control showed only moderate inhibition of tumor growth. CONCLUSION Our study has important implications for cancer immunotherapy regarding both accurate prediction of ICB response and development of better strategies to reverse existing ICB resistance. Citation Format: Guangxu Jin, Gang Xue, Rui-Sheng Wang, Ling-Yun Wu, Lance Miller, Yong Lu, Wei Zhang. High-dimensional modeling of single-cell-based CD8+ T cell exhaustion predicts response to immune checkpoint blockade [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2019; 2019 Mar 29-Apr 3; Atlanta, GA. Philadelphia (PA): AACR; Cancer Res 2019;79(13 Suppl):Abstract nr LB-218.

High-dimensional single-cell proteomics analysis identifies immune checkpoint signatures and therapeutic targets in ulcerative colitis.

Immune checkpoints are regulators of immune cells and play key roles in the modulation of immune responses. The role of checkpoints in autoimmune disease is poorly understood but likely to be central since checkpoint inhibition during cancer treatment can cause autoimmunity. We generated a high-dimensional single-cell proteomics data set from PBMCs of healthy individuals and patients with ulcerative colitis (UC) by mass cytometry, enabling systems-wide analyses of immune cell frequencies and cell type-specific expression patterns of 12 immune checkpoints. Subtle but significant changes in immune cell frequencies and checkpoint expression were observed between UC patients on different treatment regimens and between patients and healthy controls. Most strikingly, UC patients showed a reduced number of peripheral NK-cells and those cells showed an altered phenotype including increased TIGIT expression. Based on these results, we modulated NK-cell function ex vivo through targeting of TIGIT pathway members. In summary, we describe a pattern of changes in immune cell abundance and checkpoint expression as a basis for UC patient stratification and we show modulation of a corresponding immune cell subset through checkpoint targeting. Our approach can be used for the identification of pathogenic immune cell subsets and guide target selection in autoimmunity and chronic inflammation.

A 36-Dimensional Cytometry by Time of Flight (CyTOF) Analysis of De Novo Acute Myeloid Leukemia (AML) Patients Eligible for Intensive Chemotherapy

IntroductionChemotherapy induced clonal selection and development are major contributors to disease relapse and resistance to therapy in AML. An immediate response to chemotherapy in intracellular signalling networks can be detected within minutes in vivo. (Irish J. et al 2004) We hypothesize that by analyzing the immediate intracellular chemotherapy response we might reveal which cancer clones are resistant to therapy and how they might be targeted. (Anensen N. et al Clin Cancer Res 2006), (Øyan M. et al BMC Cancer 2009)CyTOF combines the principles of conventional flow cytometry with time-of-flight mass spectrometry to allow analyses of multiple parameters simultaneously in single cells. This dramatically increases the numbers of antibodies that can be applied and permits a simultaneous analysis of both intracellular signaling networks and phenotypic characterization of cells.Material and methodsWe have performed a 36- dimensional mass cytometry analysis (CyTOF), containing 21 surface markers (CD66b, CD16, CD45, CD11b, CD25, CD3, CD4, CD8, CD56, CD20, CD38, CD34, CD117, CD33, CD64, CD14, HLA-DR, Axl-1H12, CD90, CD7, CD123) and 15 intracellular markers (cCaspase-3, p-4EB-P1, p-Stat 5, p-Stat 3, p-Stat 1, p-p38, p- ERK ½, p-Akt, p-NFkB, p-CREB, p-S6, p-Axl, p-Rb, Cyclin B1, p-Histone H3) of over 40 de novo AML patients eligible for intensive chemotherapy. Peripheral blood was sampled at 4 hours and at 18- 24 hours after start of standard “3+7” induction therapy (Lowenberg B. et al Blood 2017). Bone marrow and peripheral blood from healthy donors were also included in the analysis.ResultsWell-characterized surface markers were used to perform an unsupervised clustering of the high dimensional single cell data from each AML patient before start of induction therapy (Figure 1A). This clustering resulted in 15 unique cell clusters throughout the patient cohort, illustrated by viSNE in Figure 2A. The heatmap in Figure 1A show the expression of surface markers in each cluster. Healthy cells in AML patients clustered together and were recognized by their characteristic phenotype, for example granulocytes in cluster 13 (CD66b+ and CD11b+), CD8 T-cells in cluster 1 (CD3+, CD7+ and CD8+) and monocytes in cluster 11 (CD45+, CD64+, CD38+ and HLA-DR+). Cluster 3 and 4 had high expression of the immature marker CD34 and thereby likely to represent blast populations.Intracellular signaling at time of diagnosis was then compared to 24 hours after start of induction therapy in the same populations. A lasso regression analysis identified intracellular markers that significantly correlated to patient survival, based on their change in expression between pre-treatment and at 24 hours. Figure 1C show a Kaplan-Meier curve of patient survival (24 months) between patients that had high or low expression ratio of an intracellular phosphor-specific antibody in cluster 3 and 15. Both clusters have expression of CD34 and CD117 that are frequently expressed on leukemic cells. The expression ratio of this intracellular marker was significantly correlated to survival (p<0,0001). Furthermore, these CyTOF data will be correlated to RNA-seq data, mutation status and proteomic data from the same patients.ConclusionsBy applying unsupervised clustering analysis of over 40 de novo AML patients we identified leukemic cell subpopulations that had unique patterns of intracellular phospho-specific antibodies during start of induction therapy. Changes in intracellular markers during the first 24 hours of induction therapy could significantly be related to patient survival. By analyzing molecular alterations short time after chemotherapy, we aim to reveal the sub-clonal diversity and the heterogeneous responses to treatment. Identification of cancer cell subsets that are resistant to conventional chemotherapy may in the future be used to guide therapy in AML. [Display omitted] DisclosuresGjertsen:Alden Cancer Therapy 2: Patents & Royalties: Alden Cancer Therapy II patent application in relation to CryoIT trial.; BerGenBio: Consultancy; Novartis: Consultancy; Alden Cancer Therapy 2: Membership on an entity's Board of Directors or advisory committees; Alden Cancer Therapy 2: Equity Ownership; Kinn Therapeutics: Membership on an entity's Board of Directors or advisory committees; Seattle Genetics: Consultancy; Kinn Therapeutics: Equity Ownership; Boehringer Ingelheim: Research Funding.

Deep Phenotypic Analysis Reveals a Monocyte Subpopulation Predictive of Relapse/Refractory Diffuse Large B-Cell Lymphoma

Background: Diffuse large B cell lymphoma (DLBCL) is the most frequent lymphoma subtype with up to two thirds of patients achieving cure following standard immuno-chemotherapy. However, approximately 10-15% will be primary refractory and a further 20-30% relapse. There is an urgent need for improved biomarkers to identify these poor risk patients at diagnosis who are destined to fail standard therapy so that novel approaches can be considered. The diagnostic peripheral blood monocyte count is known to be predictive of outcome in DLBCL, with higher counts identifying patients with worse outcome but mechanistic understanding is currently lacking. The aim of this study was to investigate the peripheral blood monocyte population driving this phenomenon and we present here a deep phenotypic analysis of the monocyte compartment at diagnosis in patients with extremes of outcome following immuno-chemotherapy.Methods: We identified 112 immuno-chemotherapy treated DLBCL patients with viably cryopreserved peripheral blood mononuclear cells (PBMCs) in our tissue bank. The diagnostic absolute monocyte (AMC) and lymphocyte counts (ALC) were examined as prognostic variables. For phenotypic analysis a subset of 20 patients, 10 with relapse/refractory disease within 12 months of therapy (Poor risk) and 10 with continued complete remission >24 months post therapy (Good risk), were selected. PBMCs from these patients were stained with a panel of 29 metal-tagged antibodies designed to identify immune populations and characterize the monocyte compartment and analyzed using cytometry by time-of-flight (CyTOF2TM). Samples were acquired in batches run with the same healthy donor control PBMCs to ensure consistency of staining. Normalized data were subjected to traditional Boolean gating in Cytobank (www.mrc.cytobank.org) to identify CD45+ live single cells and further analyzed using FlowSOM and CITRUS.Results: In the full 112 patient cohort, a higher diagnostic AMC (≥0.6 x 109/L) predicted worse outcome (5-year OS, 59% vs 74%, p=0.047*). The lymphocyte to monocyte ratio (LMR) was more discriminatory, with a low LMR (<2.6) predicting worse outcome (5-year OS, 49% vs 81%, p=0.0009*). CyTOF analysis of the CD45+ events from all 20 DLBCL patients were clustered in an unsupervised manner with FlowSOM and visualized according to the tSNE algorithm. This identified all expected immune populations with further subdivision into multiple clusters / sub-populations based on protein marker expression. Repeating the FlowSOM analysis by outcome revealed clusters with clear differences between the 2 groups, including striking changes to the monocyte compartment (Figure 1). CITRUS, an algorithm for automated biomarker discovery, was used to further assess the differences in the monocyte compartments between poor risk and good risk DLBCL patients. This identified a monocyte subset, representing approximately 10% of non-lymphoid CD45+ PBMCs, with increased abundance in poor risk patients with relapsed/refractory disease (p=0.0028*, Figure 2). This subset expressed markers including HLA-DR, CD11b, CD13, CD14, CD33, SIRPα (CD172a), CCR2 (CD192, monocyte chemo-attractant protein receptor), CD206 and PD-L1 (CD274).Conclusions: We confirmed the previous finding that the diagnostic AMC is predictive of outcome in our cohort of DLBCL patients and have identified differences in the peripheral blood monocyte compartment between patients with good and poor risk disease. Unsupervised analysis using FlowSOM identified monocyte clusters with greater abundance in poor risk patients. A second high-dimensional single cell data analysis algorithm, CITRUS validated this finding by revealing a stratifying monocyte sub-population that was enriched in poor risk patients. This phenotype of peripheral blood monocytes may account for the adverse nature of a higher AMC at diagnosis. We are now functionally characterizing these monocytes and evaluating their relationship to differences between the lymphoid populations of good and poor risk patients. [Display omitted] DisclosuresSeymour:Celgene: Research Funding. Gribben:Janssen: Honoraria, Research Funding; Pharmacyclics: Honoraria; Abbvie: Honoraria; Kite: Honoraria; Acerta Pharma: Honoraria, Research Funding; Wellcome Trust: Research Funding; Celgene: Consultancy, Honoraria, Research Funding; NIH: Research Funding; Unum: Equity Ownership; Roche: Honoraria; Novartis: Honoraria; Cancer Research UK: Research Funding; Medical Research Council: Research Funding; TG Therapeutics: Honoraria.

Interpretable dimensionality reduction of single cell transcriptome data with deep generative models

Single-cell RNA-sequencing has great potential to discover cell types, identify cell states, trace development lineages, and reconstruct the spatial organization of cells. However, dimension reduction to interpret structure in single-cell sequencing data remains a challenge. Existing algorithms are either not able to uncover the clustering structures in the data or lose global information such as groups of clusters that are close to each other. We present a robust statistical model, scvis, to capture and visualize the low-dimensional structures in single-cell gene expression data. Simulation results demonstrate that low-dimensional representations learned by scvis preserve both the local and global neighbor structures in the data. In addition, scvis is robust to the number of data points and learns a probabilistic parametric mapping function to add new data points to an existing embedding. We then use scvis to analyze four single-cell RNA-sequencing datasets, exemplifying interpretable two-dimensional representations of the high-dimensional single-cell RNA-sequencing data.

Model-based branching point detection in single-cell data by K-branches clustering.

MotivationThe identification of heterogeneities in cell populations by utilizing single-cell technologies such as single-cell RNA-Seq, enables inference of cellular development and lineage trees. Several methods have been proposed for such inference from high-dimensional single-cell data. They typically assign each cell to a branch in a differentiation trajectory. However, they commonly assume specific geometries such as tree-like developmental hierarchies and lack statistically sound methods to decide on the number of branching events.ResultsWe present K-Branches, a solution to the above problem by locally fitting half-lines to single-cell data, introducing a clustering algorithm similar to K-Means. These halflines are proxies for branches in the differentiation trajectory of cells. We propose a modified version of the GAP statistic for model selection, in order to decide on the number of lines that best describe the data locally. In this manner, we identify the location and number of subgroups of cells that are associated with branching events and full differentiation, respectively. We evaluate the performance of our method on single-cell RNA-Seq data describing the differentiation of myeloid progenitors during hematopoiesis, single-cell qPCR data of mouse blastocyst development, single-cell qPCR data of human myeloid monocytic leukemia and artificial data.Availability and implementationAn R implementation of K-Branches is freely available at https://github.com/theislab/kbranches.Supplementary information Supplementary data are available at Bioinformatics online.

Toward deterministic and semiautomated SPADE analysis.

SPADE stands for spanning-tree progression analysis for density-normalized events. It combines downsampling, clustering and a minimum-spanning tree to provide an intuitive visualization of high-dimensional single-cell data, which assists with the interpretation of the cellular heterogeneity underlying the data. SPADE has been widely used for analysis of high-content flow cytometry data and CyTOF data. The downsampling and clustering components of SPADE are both stochastic, which lead to stochasticity in the tree visualization it generates. Running SPADE twice on the same data may generate two different tree structures. Although they typically lead to the same biological interpretation of subpopulations present in the data, robustness of the algorithm can be improved. Another avenue of improvement is the interpretation of the SPADE tree, which involves visual inspection of multiple colored versions of the tree based on expression of measured markers. This is essentially manual gating on the SPADE tree and can benefit from automated algorithms. This article presents improvements of SPADE in both aspects above, leading to a deterministic SPADE algorithm and a software implementation for semiautomated interpretation. © 2017 International Society for Advancement of Cytometry.

Single cell proteomics in biomedicine: High-dimensional data acquisition, visualization, and analysis.

New insights on cellular heterogeneity in the last decade provoke the development of a variety of single cell omics tools at a lightning pace. The resultant high-dimensional single cell data generated by these tools require new theoretical approaches and analytical algorithms for effective visualization and interpretation. In this review, we briefly survey the state-of-the-art single cell proteomic tools with a particular focus on data acquisition and quantification, followed by an elaboration of a number of statistical and computational approaches developed to date for dissecting the high-dimensional single cell data. The underlying assumptions, unique features, and limitations of the analytical methods with the designated biological questions they seek to answer will be discussed. Particular attention will be given to those information theoretical approaches that are anchored in a set of first principles of physics and can yield detailed (and often surprising) predictions.

Cytomegalovirus Reactivation Shapes NK Cells Diversity Following Allogeneic Hematopoietic Stem Cell Transplantation

Natural killer (NK) cells are the first lymphocyte population to reconstitute following allogeneic hematopoietic stem cell transplantation (HSCT) and are key players in immune defense against viral infections and malignant transformation. NK cell numbers generally recover within the first month post-transplant, but the acquisition of mature NK cell phenotype and full functional competency can take over 6 months and is influenced by various host and donor factors. Cytomegalovirus (CMV) infection has been shown to modulate NK cell maturation after HSCT.The diversity of the NK cell repertoire is dictated by a variety of combinatorially expressed activating and inhibitory receptors that dictate the NK activation status. Moreover, whereas the expression of inhibitory receptors is primarily genetically determined, environmental factors such as viral infections influence the expression of activating receptors to a great extent.. We propose that assessment of diversity could provide a different perspective for the evaluation of the NK cell compartment after HSCT, since it is a quantitative measure that takes into account both the number and evenness of the different NK subpopulations.To better understand the factors that influence NK cell recovery after cord blood (CB) transplant (CBT) and specifically the influence of cytomegalovirus (CMV) infection on NK cell maturity, we used 40-parameter mass cytometry (CyTOF) to interrogate the NK cell repertoire. A panel including 37 monoclonal antibodies was designed to recognize NK cells lineage markers and receptors as well as intracellular markers such as transcription factors and adaptor proteins. We first evaluated and compared the diversity of NK cells in 10 CB units and peripheral blood (PB) from 20 healthy donors. We then examined the diversity of NK cells before and after CBT in 22 serially collected blood samples from in 10 CBT recipients.NK cell diversity was significantly lower in CB (mean 574, range 417-891) compared to PB samples from healthy donors (mean 3792, range 1284-5079; P=0.001), indicating less diversification within the naive CB NK compartment. After CBT, NK cell diversity was lower at earlier time point (Day30) (mean 1129, range 428-1768) compared to PB from healthy donors; P=0.01. The diversity of NK cells increased gradually over time following CBT (Day 30 mean 1129 range 428-1768; Day 60, mean 1185, range 515-1864; 4 months, mean 1711 range 597-2640).We also compared the diversity of NK cells in the PB of healthy CMV seronegative (n=10) and seropositive adult donors (n=10). The diversity of NK cells was higher in CMV seropositive vs. CMV seronegative healthy donors (3887 vs 2473; P=0.04). This difference in NK diversity was even more pronounced within the KIR positive (mean 1701, range, 981-2152) compared to the KIR negative subset (mean 551, range 456-647; P=0.02), indicating that CMV infection increases the richness of mature NK cells. In keeping with these findings, CMV infection after CBT was associated with a significantly greater diversity of NK cells, especially within the KIR positive compartment (mean 604, range 207-1035) compared to the KIR negative subset (mean 283, 257-457; P=0.025). However, in CMV negative patients, we found no difference in diversity within the KIR positive and negative subsets (mean 1120 vs. 1366; P=0.28). Taken together, these data suggest that NK cell diversity reflects NK cells differentiation and maturation, and that CMV shapes NK cell diversity, especially within the KIR positive compartment.To further understand how CMV influences NK cells diversity, we examined the top 15 NK cell subsets and their distribution at multiple timepoints before and after CMV reactivation post-CBT. CMV infection post-CBT was associated with a significant change in the distribution of NK subsets within the KIR positive population, with the top 15 subsets prior to CMV reactivation being mostly replaced by the emergence of new subsets. In contrast, the top 15 subsets within the KIR negative NK population remained stable. These data suggest that CMV drives NK cell maturation by differentiating KIR positive NK cells.In summary, we used high-dimensional single-cell data to evaluate NK cell reconstitution following HSCT. These data can help us better understand the biology of NK cell recovery after HSCT and discover the functional significance of NK cell diversity in the setting of viral infections. DisclosuresChamplin:Ziopharm Oncology: Equity Ownership, Patents & Royalties; Intrexon: Equity Ownership, Patents & Royalties.

SCALABLE VISUALIZATION FOR HIGH-DIMENSIONAL SINGLE-CELL DATA.

Single-cell analysis can uncover the mysteries in the state of individual cells and enable us to construct new models about the analysis of heterogeneous tissues. State-of-the-art technologies for single-cell analysis have been developed to measure the properties of single-cells and detect hidden information. They are able to provide the measurements of dozens of features simultaneously in each cell. However, due to the high-dimensionality, heterogeneous complexity and sheer enormity of single-cell data, its interpretation is challenging. Thus, new methods to overcome high-dimensionality are necessary. Here, we present a computational tool that allows efficient visualization of high-dimensional single-cell data onto a low-dimensional (2D or 3D) space while preserving the similarity structure between single-cells. We first construct a network that can represent the similarity structure between the high-dimensional representations of single-cells, and then, embed this network into a low-dimensional space through an efficient online optimization method based on the idea of negative sampling. Using this approach, we can preserve the high-dimensional structure of single-cell data in an embedded low-dimensional space that facilitates visual analyses of the data.

Visualization and cellular hierarchy inference of single-cell data using SPADE.

High-throughput single-cell technologies provide an unprecedented view into cellular heterogeneity, yet they pose new challenges in data analysis and interpretation. In this protocol, we describe the use of Spanning-tree Progression Analysis of Density-normalized Events (SPADE), a density-based algorithm for visualizing single-cell data and enabling cellular hierarchy inference among subpopulations of similar cells. It was initially developed for flow and mass cytometry single-cell data. We describe SPADE's implementation and application using an open-source R package that runs on Mac OS X, Linux and Windows systems. A typical SPADE analysis on a 2.27-GHz processor laptop takes ∼5 min. We demonstrate the applicability of SPADE to single-cell RNA-seq data. We compare SPADE with recently developed single-cell visualization approaches based on the t-distribution stochastic neighborhood embedding (t-SNE) algorithm. We contrast the implementation and outputs of these methods for normal and malignant hematopoietic cells analyzed by mass cytometry and provide recommendations for appropriate use. Finally, we provide an integrative strategy that combines the strengths of t-SNE and SPADE to infer cellular hierarchy from high-dimensional single-cell data.

Robust lineage reconstruction from high-dimensional single-cell data.

Single-cell gene expression data provide invaluable resources for systematic characterization of cellular hierarchy in multi-cellular organisms. However, cell lineage reconstruction is still often associated with significant uncertainty due to technological constraints. Such uncertainties have not been taken into account in current methods. We present ECLAIR (Ensemble Cell Lineage Analysis with Improved Robustness), a novel computational method for the statistical inference of cell lineage relationships from single-cell gene expression data. ECLAIR uses an ensemble approach to improve the robustness of lineage predictions, and provides a quantitative estimate of the uncertainty of lineage branchings. We show that the application of ECLAIR to published datasets successfully reconstructs known lineage relationships and significantly improves the robustness of predictions. ECLAIR is a powerful bioinformatics tool for single-cell data analysis. It can be used for robust lineage reconstruction with quantitative estimate of prediction accuracy.