Comparison of five clustering algorithms to classify phytoplankton from flow cytometry data.

BACKGROUND: Artificial neural networks (ANNs) have been shown to be valuable in the analysis of analytical flow cytometric (AFC) data in aquatic ecology. Automated extraction of clusters is an important first stage in deriving ANN training data from field samples, but AFC data pose a number of challenges for many types of clustering algorithm. The fuzzy k-means algorithm recently has been extended to address nonspherical clusters with the use of scatter matrices. Four variants were proposed, each optimizing a different measure of clustering "goodness." METHODS: With AFC data obtained from marine phytoplankton species in culture, the four fuzzy k-means algorithm variants were compared with each other and with another multivariate clustering algorithm based on critical distances currently used in flow cytometry. RESULTS: One of the algorithm variants (adaptive distances, also known as the Gustafson--Kessel algorithm) was found to be robust and reliable, whereas the others showed various problems. CONCLUSIONS: The adaptive distances algorithm was superior in use to the clustering algorithms against which it was tested, but the problem of automatic determination of the number of clusters remains to be addressed.     

Hierarchical clustering of flow cytometry data for the study of conventional central chondrosarcoma

We have investigated the use of hierarchical clustering of flow cytometry data to classify samples of conventional central chondrosarcoma, a malignant cartilage forming tumor of uncertain cellular origin, according to similarities with surface marker profiles of several known cell types. Human primary chondrosarcoma cells, articular chondrocytes, mesenchymal stem cells, fibroblasts, and a panel of tumor cell lines from chondrocytic or epithelial origin were clustered based on the expression profile of eleven surface markers. For clustering, eight hierarchical clustering algorithms, three distance metrics, as well as several approaches for data preprocessing, including multivariate outlier detection, logarithmic transformation, and z‐score normalization, were systematically evaluated. By selecting clustering approaches shown to give reproducible results for cluster recovery of known cell types, primary conventional central chondrosacoma cells could be grouped in two main clusters with distinctive marker expression signatures: one group clustering together with mesenchymal stem cells (CD49b‐high/CD10‐low/CD221‐high) and a second group clustering close to fibroblasts (CD49b‐low/CD10‐high/CD221‐low). Hierarchical clustering also revealed substantial differences between primary conventional central chondrosarcoma cells and established chondrosarcoma cell lines, with the latter not only segregating apart from primary tumor cells and normal tissue cells, but clustering together with cell lines from epithelial lineage. Our study provides a foundation for the use of hierarchical clustering applied to flow cytometry data as a powerful tool to classify samples according to marker expression patterns, which could lead to uncover new cancer subtypes. J. Cell. Physiol. 225: 601–611, 2010. © 2010 Wiley‐Liss, Inc.     

Nonparametric discriminant analysis of phytoplankton species using data from analytical flow cytometry

BACKGROUND: Analytical flow cytometry (AFC) provides rapid and accurate measurement of particles from heterogeneous populations. AFC has been used to classify and identify phytoplankton species, but most methods of discriminant analysis of resulting data have depended on normality assumptions and outcomes have been disappointing. METHODS AND RESULTS: In this study, we consider nonparametric methods based on density estimation. In addition to the familiar kernel method, methods based on wavelets are also implemented. Full five-dimensional wavelet estimation proves to be computationally prohibitive with current workstation power, so we employ projection pursuit for reduction of dimensionality. AFC typically produces very large samples, so we also investigate data simplification through binning. Further modifications to the discrimination strategy are suggested by specific features of phytoplankton data, namely, a hierarchical group structure, the possible presence of many groups, and the likelihood of encountering an aberrant group in a test sample. CONCLUSIONS: We apply all the resultant procedures to appropriate subsets of a very large data set, demonstrate their efficacy, and compare their error rates with those of more conventional methods. We further show that incorporation of the specific features of phytoplankton data into the analysis leads to improved results and provides a general framework for analysis of such data.     

Combination of automated high throughput platforms, flow cytometry, and hierarchical clustering to detect cell state.

BACKGROUND: This study examined whether hierarchical clustering could be used to detect cell states induced by treatment combinations that were generated through automation and high-throughput (HT) technology. Data-mining techniques were used to analyze the large experimental data sets to determine whether nonlinear, non-obvious responses could be extracted from the data. METHODS: Unary, binary, and ternary combinations of pharmacological factors (examples of stimuli) were used to induce differentiation of HL-60 cells using a HT automated approach. Cell profiles were analyzed by incorporating hierarchical clustering methods on data collected by flow cytometry. Data-mining techniques were used to explore the combinatorial space for nonlinear, unexpected events. Additional small-scale, follow-up experiments were performed on cellular profiles of interest. RESULTS: Multiple, distinct cellular profiles were detected using hierarchical clustering of expressed cell-surface antigens. Data-mining of this large, complex data set retrieved cases of both factor dominance and cooperativity, as well as atypical cellular profiles. Follow-up experiments found that treatment combinations producing "atypical cell types" made those cells more susceptible to apoptosis. CONCLUSIONS Hierarchical clustering and other data-mining techniques were applied to analyze large data sets from HT flow cytometry. From each sample, the data set was filtered and used to define discrete, usable states that were then related back to their original formulations. Analysis of resultant cell populations induced by a multitude of treatments identified unexpected phenotypes and nonlinear response profiles.     

Raman microspectroscopic discrimination of TCam-2 cultures reveals the presence of two sub-populations of cells

TCam-2 cells are the main in vitro model for investigations into seminomatous tumors. However, despite their widespread use, questions remain regarding the cells’ homogeneity and consequently how representative they are of seminomas. We assess the TCam-2 cell line using routine and novel authentication methods to determine its homogeneity, identify any cellular sub-populations and resolve whether any changes could be due to generational differentiation. TCam-2, embryonal carcinoma cells (2102EP) and breast cancer cell (MCF7) lines were assessed using qRT-PCR, immunocytochemistry, flow cytometry and short tandem repeat analyses. Raman maps of individual cells (minimum of 10) and single scan spectra from 200 cells per culture were obtained. TCam-2s displayed the characteristic marker gene expression pattern for seminoma, were uniform in size and granularity and short tandem repeat analysis showed no contamination. However, based only on physical parameters, flowcytometry was unable to differentiate between TCam-2 and 2102EPs. Raman maps of TCam-2s comprised three equally distributed, distinct spectral patterns displaying large intercellular single spectral variation. All other cells showed little variation. Principal component, cluster and local spectral angle analyses indicated that the TCam-2s contained two different types of cells, one of which comprised two subgroups and was similar to some 2102EP cells. Protein expression corroborated the presence of different cells and generational differences. The detailed characterization provided by the Raman spectra, augmented by the routine methods, provide substantiation to the long-held suspicion that TCam-2 are not homogeneous but comprise differing cell populations, one of which may be embryonal carcinoma in origin.     

Alveolar macrophage-environmental particle interaction: analysis by flow cytometry

Inhaled particulates such as pollutant particles, allergens, and microorganisms are rapidly cleared by alveolar macrophages (AMs). Methods for analysis of AM-particle interaction have been hindered by the lack of a convenient assay. Flow cytometry offers rapid, sensitive, and reproducible measurements of single cells in suspension. Multiple parameters can be measured in real time. Here we will review the application of flow cytometry to the study and characterization of AM receptors for unopsonized environmental particles. We will discuss the role of this technique in identifying a key AM receptor system involved in lung defense. Multiparametric flow cytometry to analyze intracellular functional parameters, though a powerful and unique tool, needs to be interpreted with caution. We will also discuss the advantages and limitations of flow cytometry in analysis of AM-particle interaction.     

Quantitative fluorescence cytometric analysis of Bcl-2 levels in tumor cells exhibiting a wide range of inherent Bcl-2 protein expression: correlation with Western blot analysis

BACKGROUND: A protocol to measure a wide range of Bcl-2 protein expression using quantitative fluorescence cytometry (QFCM) in different cell types was developed for use with flow cytometry. Bcl-2 measurements obtained by flow cytometry were correlated with Western blot Bcl-2 measurements to confirm specificity of the Bcl-2-FITC staining. This protocol was applied to measure absolute levels of Bcl-2 protein in different tumor cell lines including Bcl-2-transfected breast carcinoma cell lines and in peripheral blood lymphocytes (PBL). METHODS: HL-60, K562, DOHH2, Jurkat, MDA435/LCC6, MCF7 cell lines, and PBL derived from normal donors were fixed, permeabilized, stained with anti-Bcl-2-FITC antibody and evaluated by QFCM. In parallel, the same cells were evaluated for Bcl-2 protein expression by Western blot analysis. Mitochondrial localization of anti-Bcl-2-FITC antibody inside cells was confirmed using fluorescence imaging microscopy. RESULTS: Bcl-2 expression in different cell types could be accurately quantified based on antibody-binding capacity (ABC) ranging from 12.6 x 10(3) antibody-binding sites in HL-60 cells to 1.64 x 10(6) antibody-binding sites in a Bcl-2-transfected MDA435/LCC6 clone. The data from flow cytometry analysis correlated well with Western analysis (R(2) = 0.78). Bcl-2-FITC staining colocalized with dyes specific for mitochondria. CONCLUSIONS: The Bcl-2 staining protocol described here was shown to be specific, sensitive, and it was able to provide higher resolution as well as more reproducible quantitation of Bcl-2 protein content in cells when compared with Western blot methods. Quantitation of Bcl-2 content in cells by QFCM may be useful for monitoring Bcl-2 expression in cells undergoing various treatments in vitro and in vivo.     

Differences in the flow and absorption cytometric DNA distributions of mouse hepatocytes and tumor cells

We have determined the DNA content, the ploidy levels, and the percentages of different cell types present in small and large mouse mammary tumors as well as in young and old mouse livers by using absorption and flow cytometry. Absorption cytometry data indicated a significant increase in the proportion of transformed G0/G1 cells in the tumors as compared to that of the stromal G0/G1 cells with progressive tumor growth. This increase was not detected by flow cytometry. In both young and old mouse livers, a small number of cells of higher ploidy (8C and 16C) were detected by absorption cytometry but were not apparent in histograms obtained by flow cytometry. Furthermore, changes in the proportions of liver cells of different ploidy with age were apparent in absorption cytometry data but not in flow cytometry data. In one mouse liver experiment, a 6C cell peak appeared in the flow cytometry histogram, but a direct measurement of DNA content by absorption cytometry failed to detect cells with such a peak. We therefore believe that some caution may be warranted in the use of flow cytometry alone for evaluation of DNA distributions and of the proportions of different types of cells in complex solid tissues.     

Detection and monitoring of normal and leukemic cell populations with hierarchical clustering of flow cytometry data

Flow cytometry is a valuable tool in research and diagnostics including minimal residual disease (MRD) monitoring of hematologic malignancies. However, its gradual advancement toward increasing numbers of fluorescent parameters leads to information rich datasets, which are challenging to analyze by standard gating and do not reflect the multidimensionality of the data. We have developed a novel method to analyze complex flow cytometry data, based on hierarchical clustering analysis (HCA) but with a new underlying algorithm, using Mahalanobis distance measure. HCA is scalable to analyze complex multiparameter datasets (here demonstrated on up to 12 color flow cytometry and on a 20-parameter synthetic dataset). We have validated this method by comparison with standard gating approaches when performed independently by expert cytometrists. Acute lymphoblastic leukemia blast populations were analyzed in diagnostic and follow-up datasets (n = 123) from three centers. HCA results correlated very well (Passing-Bablok correlation coefficient = 0.992, slope = 1, intercept = -0.01) with standard gating data obtained by the I-BFM FLOW-MRD study group. To further improve the performance in follow-up samples with low MRD levels and to automate MRD detection, we combined HCA with support vector machine (SVM) learning. HCA in combination with SVM provides a novel diagnostic tool that not only allows analysis of increasingly complex flow cytometry data but also is less observer-dependent compared with classical gating and has potential for automation.     

Polychromatic flow cytometry: a rapid method for the reduction and analysis of complex multiparameter data.

BACKGROUND: Recent advances in flow cytometry have resulted in the development of reliable techniques for performing polychromatic (5-17 color) flow cytometry analysis. However, the data reduction and analysis involved in the resolution of hundreds of possible cellular subphenotypes identified, using a single polychromatic flow cytometry staining panel, presents a major obstacle to the successful application of this technology. METHODS: To generate two distinct collections of T cell populations with differentially expressed surface markers, cryopreserved lymph node cells from 5 melanoma patients vaccinated with the modified gp100(209-2M) melanoma peptide were stimulated with cognate peptide and cultured in either IL-21 + low-dose IL-2 or IL-15 + low-dose IL-2. In vitro stimulated (IVS) cells were interrogated using 8-color flow cytometry. Data were analyzed using Winlist Hyperlog and FCOM software, and 32 T cell subsets were resolved for each culture condition. Hierarchical clustering analysis was applied to the relative percentages of each subphenotype for both IVS conditions to determine if unique cell surface marker expression signatures were produced for each IVS culture. RESULTS: Sequential data analysis using Hyperlog and FCOM demonstrated that lymphocytes cultured in IL-21 + IL-2 had a distinctively different set of subphenotype signatures compared to cells grown in IL-15 + IL-2 for all 5 patients. Importantly, subsequent cluster analysis of all 32 subphenotype frequencies in each IVS test condition for all 5 patients reproducibly demonstrated that cellular subphenotypes produced after IL-21 + IL-2 IVS partitioned separately from subphenotypes produced by IL-15 + IL-2 IVS. CONCLUSIONS: The integrated sequential use of Hyperlog and FCOM software with cluster analysis algorithms for the reduction and analysis of polychromatic flow cytometry data produces an effective, rapid technique for the assessment of complex patterns of subphenotype expression between and within multiple test samples. This approach to data analysis may enhance the use of polychromatic flow cytometry for both research and clinical applications.     

Total nitrogen and ammonia removal prediction in horizontal subsurface flow constructed wetlands: use of artificial neural networks and development of a design equation

The aim of this paper is to examine if artificial neural networks (ANNs) can predict nitrogen removal in horizontal subsurface flow (HSF) constructed wetlands (CWs). ANN development was based on experimental data from five pilot-scale CW units. The proper selection of the components entering the ANN was achieved using principal component analysis (PCA), which identified the main factors affecting TN removal, i.e., porous media porosity, wastewater temperature and hydraulic residence time. Two neural networks were examined: the first included only the three factors selected from the PCA, and the second included in addition meteorological parameters (i.e., barometric pressure, rainfall, wind speed, solar radiation and humidity). The first model could predict TN removal rather satisfactorily (R(2)=0.53), and the second resulted in even better predictions (R(2)=0.69). From the application of the ANNs, a design equation was derived for TN removal prediction, resulting in predictions comparable to those of the ANNs (R(2)=0.47). For the validation of the results of the ANNs and of the design equation, available data from the literature were used and showed a rather satisfactory performance.     

Extracting a cellular hierarchy from high-dimensional cytometry data with SPADE

The ability to analyze multiple single-cell parameters is critical for understanding cellular heterogeneity. Despite recent advances in measurement technology, methods for analyzing high-dimensional single-cell data are often subjective, labor intensive and require prior knowledge of the biological system. To objectively uncover cellular heterogeneity from single-cell measurements, we present a versatile computational approach, spanning-tree progression analysis of density-normalized events (SPADE). We applied SPADE to flow cytometry data of mouse bone marrow and to mass cytometry data of human bone marrow. In both cases, SPADE organized cells in a hierarchy of related phenotypes that partially recapitulated well-described patterns of hematopoiesis. We demonstrate that SPADE is robust to measurement noise and to the choice of cellular markers. SPADE facilitates the analysis of cellular heterogeneity, the identification of cell types and comparison of functional markers in response to perturbations.     

Chlortetracycline analysis of boar spermatozoa after incubation,flow cytometric sorting,cooling, or cryopreservation

In this study, the effects of staining procedure with chlortetracycline (CTC) and method of analysis of boar spermatozoa after staining were examined. The hypothesis that incubation, flow cytometric sorting, cooling, and cryopreservation cause changes to boar sperm membranes which resemble capacitation and the acrosome reaction was also tested. Membrane status was evaluated by flow cytometry and by fluorescence microscopy after staining with CTC, and acrosome integrity was checked by flow cytometry after staining with FITC-pisum sativum agglutenin and propidium iodide (PI). Flow cytometry was also used to assess viability (percentages of live and dead cells) of boar sperm after staining with SYBR-14 and PI. Staining of spermatozoa with CTC alone and in combination with PI and/or Hoechst 33342 had no effect on the proportion of spermatozoa allocated to the F (uncapacitated), B (capacitated), or AR (acrosome-reacted) CTC fluorescent staining categories. The mean percentages of acrosome-intact and acrosome-reacted cells were 88.4 and 6.8 or 0.8 and 96.5 in semen treated with 0 or 100 μg/ml lysophosphatidylchloine (LPC), respectively (P < 0.001). Most spermatozoa were also in the AR CTC-stained category after treatment with LPC compared with a small proportion in the controls. Using flow cytometry to examine sperm suspensions stained with CTC, a gated population of spermatozoa with low fluorescence (population 1) comprised predominantly F-pattern cells (F-pattern: population 1 vs. population 2, 80.5 vs. 14.4%; P < 0.001), whereas population 2 (high fluorescence) comprised mainly B-pattern cells (B-pattern: population 1 vs. population 2, 8.5 vs. 62.3%; P < 0.001). Incubation (38°C, 4 hr), flow sorting, cooling (to 15 or 5°C) and freezing reduced the proportion of F-pattern and live spermatozoa, and increased the proportion of B-, AR-pattern, and dead spermatozoa, in comparison with fresh semen. There were more membrane changes in spermatozoa cooled to 5°C (30.4, 48.5, 21.1%) than in those cooled to 15°C (56.1, 32.6, 11.5% F-, B-, and AR-pattern spermatozoa, respectively). Mol. Reprod. Dev. 46:408–418, 1997. © 1997 Wiley-Liss, Inc.     

Variable selection and multivariate methods for the identification of microorganisms by flow cytometry.

BACKGROUND: When exploited fully, flow cytometry can be used to provide multiparametric data for each cell in the sample of interest. While this makes flow cytometry a powerful technique for discriminating between different cell types, the data can be difficult to interpret. Traditionally, dual-parameter plots are used to visualize flow cytometric data, and for a data set consisting of seven parameters, one should examine 21 of these plots. A more efficient method is to reduce the dimensionality of the data (e.g., using unsupervised methods such as principal components analysis) so that fewer graphs need to be examined, or to use supervised multivariate data analysis methods to give a prediction of the identity of the analyzed particles. MATERIALS AND METHODS: We collected multiparametric data sets for microbiological samples stained with six cocktails of fluorescent stains. Multivariate data analysis methods were explored as a means of microbial detection and identification. RESULTS: We show that while all cocktails and all methods gave good accuracy of predictions (>94%), careful selection of both the stains and the analysis method could improve this figure (to > 99% accuracy), even in a data set that was not used in the formation of the supervised multivariate calibration model. CONCLUSIONS: Flow cytometry provides a rapid method of obtaining multiparametric data for distinguishing between microorganisms. Multivariate data analysis methods have an important role to play in extracting the information from the data obtained. Artificial neural networks proved to be the most suitable method of data analysis.     

Photothermal flow cytometry in vitro for detection and imaging of individual moving cells.

BACKGROUND: Photothermal (PT) cytometry has recently demonstrated great potential for the label-free detection of nonfluorescent cells under static conditions. The goal of our investigation was to expand this technique to the detection of flowing cells in vitro. METHODS: Cells in flow were irradiated with short, tunable laser pulses (420-2,300 nm, 8 ns), and the absorbed energy was detected by monitoring of the temperature-dependent variations in the refractive index in the cells with a second, collinear probe beam in two modes: (a) PT imaging of single cells with a pulsed probe beam (639 nm, 13 ns) and (b) thermolens monitoring of the integral PT responses from individual cells as whole with a continuous-wave probe beam (633 nm, 2 mW). RESULTS: PT flow cytometry at the current speed of analysis of 10 cell/s, with the capability to image selected cells of interest flowing at velocities up to 2 m/s, demonstrated the capability for (a) label-free detection of flowing single cells (e.g., blood and cancer cells) on the basis of the differences in their endogenous absorption properties, (b) identification of cells labeled with gold nanoparticles, (c) rapid cell viability testing, (d) aggregation immunoassay, and (e) optimization of selective nanophotothermolysis. CONCLUSIONS: PT cytometry can be extended to the study of cells in flow. This new technique increases the speed of cell analysis approximately 10(2) times over that of conventional PT technique, with the potential to achieve a rate of 10(4)-10(5) cells/s in specific PT applications, which has previously been realized only with cells under static conditions.     

DIFFERENTIATION OF ANTIBODY-FORMING CELLS IN TOAD SPLEEN : A Study Using Density and Sedimentation Velocity Cell Separation

Antibody-forming cells (AFC), developing in toad spleen after stimulation with polymerized flagellin, were studied with an immune adherence assay. Differentiation was followed by several parameters: thymidine uptake to monitor dividing cells; equilibrium density centrifugation in albumin gradients to monitor cell density; microscopic measurements and sedimentation velocity separation to monitor cell size; stained preparations to follow cell morphology. Almost all AFC observed early in the response were dividing cells; the proportion of dividing AFC dropped to 4% 2 wk after stimulation. The earliest AFC detected (3 days) formed a relatively homogeneous light density population, and were purified 17-fold by equilibrium density centrifugation. As the response developed, additional denser peaks were found, so that late in the response dense AFC predominated. Dividing AFC were confined to the light density region throughout the response. Cell diameter measurements revealed that the earliest AFC were all very large cells. In a manner analogous to the density changes, smaller AFC appeared as the response developed until they finally comprised the majority of the AFC population. Dividing AFC were always relatively large, but encompassed a wide range of sizes. Sedimentation velocity separation was employed in a closer study of the immature AFC; they were purified 140-fold by this procedure. The earliest AFC consisted of several readily separable size populations in the range 9–18 µ diameter. The presence of separate peaks related by factors of two in volume suggested that the largest cells undergo a series of halving divisions before entering a division growth cycle. The results suggest an AFC differentiation sequence from a very large, light density, dividing "blast" cell to a nondividing cell with the size, density, and morphological appearance of a small lymphocyte. Stages of this sequence can be defined and selected out for investigation, using sedimentation velocity and equilibrium density centrifugation as complementary techniques.     

Binding of fluoresceinated epidermal growth factor to A431 cell sub-populations studied using a model-independent analysis of flow cytometric fluorescence data.

A method is developed for determining ligand-cell association parameters from a model-free analysis of data obtained with a flow cytometer. The method requires measurement of the average fluorescence per cell as a function of ligand and cell concentration. The analysis is applied to data obtained for the binding of fluoresceinated epidermal growth factor to a human epidermoid carcinoma cell line, A431. The results indicate that the growth factor binds to two classes of sites on A431 cells: 4 X 10(4) sites with a dissociation constant (KD) of less than or equal to 20 pM, and 1.5 X 10(6) sites with a KD of 3.7 nM. A derived plot of the average fluorescence per cell versus the average number of bound ligands per cell is used to construct binding isotherms for four sub-populations of A431 cells fractionated on the basis of low-angle light scatter. The four sub-populations bind the ligand with equal affinity but differ substantially in terms of the number of binding sites per cell. We also use this new analysis to critically evaluate the use of 'Fluorotrol' as a calibration standard in flow cytometry.     

Determination of critical epitope of PcMab-47 against human podocalyxin

Podocalyxin (PODXL) is a type I transmembrane protein, which is highly glycosylated. PODXL is expressed in some types of human cancer tissues including oral, breast, and lung cancer tissues and may promote tumor growth, invasion, and metastasis. We previously produced PcMab-47, a novel anti-PODXL monoclonal antibody (mAb) which reacts with endogenous PODXL-expressing cancer cell lines and normal cells independently of glycosylation in Western blot, flow cytometry, and immunohistochemical analysis. In this study, we used enzyme-linked immunosorbent assay (ELISA), flow cytometry, and immunohistochemical analysis to determine the epitope of PcMab-47. The minimum epitope of PcMab-47 was found to be Asp207, His208, Leu209, and Met210. A blocking peptide containing this minimum epitope completely neutralized PcMab-47 reaction against oral cancer cells by flow cytometry and immunohistochemical analysis. These findings could lead to the production of more functional anti-PODXL mAbs, which are advantageous for antitumor activities.     

An introduction to mass cytometry: fundamentals and applications

Mass cytometry addresses the analytical challenges of polychromatic flow cytometry by using metal atoms as tags rather than fluorophores and atomic mass spectrometry as the detector rather than photon optics. The many available enriched stable isotopes of the transition elements can provide up to 100 distinguishable reporting tags, which can be measured simultaneously because of the essential independence of detection provided by the mass spectrometer. We discuss the adaptation of traditional inductively coupled plasma mass spectrometry to cytometry applications. We focus on the generation of cytometry-compatible data and on approaches to unsupervised multivariate clustering analysis. Finally, we provide a high-level review of some recent benchmark reports that highlight the potential for massively multi-parameter mass cytometry.