High-Throughput Analysis of Clinical Flow Cytometry Data by Automated Gating

Using Artificial Intelligence to Interpret Clinical Flow Cytometry Datasets for Automated Disease Diagnosis and/or Monitoring

Flow cytometry (FC) is routinely used for hematological disease diagnosis and monitoring. Advancement in this technology allows us to measure an increasing number of markers simultaneously, generating complex high-dimensional datasets. However, current analytic software and methods rely on experienced analysts to perform labor-intensive manual inspection and interpretation on a series of 2-dimensional plots via a complex, sequential gating process. With an aggravating shortage of professionals and growing demands, it is very challenging to provide the FC analysis results in a fast, accurate, and reproducible way. Artificial intelligence has been widely used in many sectors to develop automated detection or classification tools. Here we describe a type of machine learning method for developing automated disease classification and residual disease monitoring on clinical flow datasets.

flowSim: Near duplicate detection for flow cytometry data

The analysis of large amounts of data is important for the development of machine learning (ML) models. flowSim is the first algorithm designed to visualize, detect and remove highly redundant information in flow cytometry (FCM) training sets to decrease the computational time for training and increase the performance of ML algorithms by reducing overfitting. flowSim performs near duplicate image detection by combining community detection algorithms with the density analysis of the marker expression values. flowSim clustering compared to consensus manual clustering on a dataset composed of 160 images of bivariate FCM data had a mean Adjusted Rand Index of 0.90, demonstrating its efficiency in identifying similar patterns. flowSim selectively discarded near duplicate files in datasets constructed with known redundancy, and removed 92.6% of FCM images in a dataset of over 500,000 drawn from public repositories.

Full spectrum flow cytometry and mass cytometry: A 32-marker panel comparison

High-dimensional single-cell data has become an important tool in unraveling the complexity of the immune system and its involvement in homeostasis and a large array of pathologies. As technological tools are developed, researchers are adopting them to answer increasingly complex biological questions. Up until recently, mass cytometry (MC) has been the main technology employed in cytometric assays requiring more than 29 markers. Recently, however, with the introduction of full spectrum flow cytometry (FSFC), it has become possible to break the fluorescence barrier and go beyond 29 fluorescent parameters. In this study, in collaboration with the Stanford Human Immune Monitoring Center (HIMC), we compared five patient samples using an established immune panel developed by the HIMC using their MC platform. Using split samples and the same antibody panel, we were able to demonstrate highly comparable results between the two technologies using multiple data analysis approaches. We report here a direct comparison of two technology platforms (MC and FSFC) using a 32-marker flow cytometric immune monitoring panel that can identify all the previously described and anticipated immune subpopulations defined by this panel.

The role of CD8+ Granzyme B+ T cells in the pathogenesis of Takayasu's arteritis

OBJECTIVE

T cell-mediated immune response plays a key role in Takayasu arteritis (TAK). Although previous studies have showed the roles of CD4⁺T cell and its subsets in TAK, the change of CD8⁺ T cell subsets remains unclear. This study investigated the role of CD8⁺ T cell subsets in TAK.

METHODS

The study consisted of 56 TA patients and 51 healthy controls. The percentages of CD8⁺T cells, CD8⁺GranzymeB⁺ T cells, CD8⁺Perforin⁺ T cells, and CD8⁺IFN-γ⁺ T cells in blood samples were analyzed by flow cytometry.

RESULTS

We found that the percentages of CD8⁺GranzymeB⁺ T cells (P = 0.030), CD8⁺Perforin⁺ T cells (P = 0.000), and CD8⁺IFN-γ⁺ T cells (P = 0.002) in CD8⁺T cells were higher in TAK patients compared to control group. After 6 months of treatment, the proportion of CD8⁺T cells in lymphocytes were significantly lower in TAK patients than the baseline assessment (P = 0.033). A lower ratio of CD8⁺GranzymeB⁺ T cells/CD8⁺ T cells were showed in TAK patents after treatment compared with TAK patients before treatments (P = 0.011). The change of CD8⁺GranzymeB⁺ T cells/CD8⁺ T cells ratio was positively correlated with the change of ITAS (r = 0.721, P = 0.002) and ITAS-A (r = 0.637, P = 0.008). Finally, the immunofluorescence staining showed the infiltration of CD8⁺ Granzyme B ⁺ cells in the aortic tissue of TAK patients.

CONCLUSION

Our results disclose that the CD8⁺ T lymphocytes may play a role in TAK pathogenesis. Targeting CD8⁺GranzymeB⁺ T lymphocytes or Granzyme B inhibitors could be a potential therapeutic approach for the treatment of TAK. Key Points • Our study investigated role the of CD8⁺ T cell subsets in TAK. • We found the percentages of CD8⁺GranzymeB⁺ T cells, CD8⁺Perforin⁺ T cells, and CD8⁺IFN-γ⁺ T cells in CD3⁺CD8⁺T cells were higher in TAK patients. • The proportion of CD8⁺T cells in lymphocytes and the ratio of CD8⁺GranzymeB⁺ T cells/CD8⁺ T cells were significantly lower in TAK patients after treatment.

Automated analysis of bacterial flow cytometry data with FlowGateNIST

Flow cytometry is commonly used to evaluate the performance of engineered bacteria. With increasing use of high-throughput experimental methods, there is a need for automated analysis methods for flow cytometry data. Here, we describe FlowGateNIST, a Python package for automated analysis of bacterial flow cytometry data. The main components of FlowGateNIST perform automatic gating to differentiate between cells and background events and then between singlet and multiplet events. FlowGateNIST also includes a method for automatic calibration of fluorescence signals using fluorescence calibration beads. FlowGateNIST is open source and freely available with tutorials and example data to facilitate adoption by users with minimal programming experience.

Case-control diagnostic accuracy study of a non-sputum CD38-based TAM-TB test from a single milliliter of blood

CD4 T cell phenotyping-based blood assays have the potential to meet WHO target product profiles (TPP) of non-sputum-biomarker-based tests to diagnose tuberculosis (TB). Yet, substantial refinements are required to allow their implementation in clinical settings. This study assessed the real time performance of a simplified T cell activation marker (TAM)-TB assay to detect TB in adults from one millilitre of blood with a 24 h turnaround time. We recruited 479 GeneXpert positive cases and 108 symptomatic but GeneXpert negative controls from presumptive adult TB patients in the Temeke District of Dar-es-Salaam, Tanzania. TAM-TB assay accuracy was assessed by comparison with a composite reference standard comprising GeneXpert and solid culture. A single millilitre of fresh blood was processed to measure expression of CD38 or CD27 by CD4 T cells producing IFN-γ and/or TNF-α in response to a synthetic peptide pool covering the sequences of Mycobacterium tuberculosis (Mtb) ESAT-6, CFP-10 and TB10.4 antigens on a 4-color FACSCalibur apparatus. Significantly superior to CD27 in accurately diagnosing TB, the CD38-based TAM-TB assay specificity reached 93.4% for a sensitivity of 82.2% with an area under the receiver operating characteristics curve of 0.87 (95% CI 0.84-0.91). The assay performance was not significantly affected by HIV status. To conclude, we successfully implemented TAM-TB immunoassay routine testing with a 24 h turnaround time at district level in a resource limited setting. Starting from one millilitre of fresh blood and being not influenced by HIV status, TAM-TB assay format and performance appears closely compatible with the optimal TPP accuracy criteria defined by WHO for a non-sputum confirmatory TB test.

CyAnno: A semi-automated approach for cell type annotation of mass cytometry datasets

MOTIVATION

For immune system monitoring in large-scale studies at the single-cell resolution using CyTOF, (semi-)automated computational methods are applied for annotating live cells of mixed cell types. Here, we show that the live cell pool can be highly enriched with undefined heterogeneous cells, i.e., 'ungated' cells, and that current semi-automated approaches ignore their modeling resulting in misclassified annotations.

RESULT

We introduce 'CyAnno', a novel semi-automated approach for deconvoluting the unlabeled cytometry dataset based on a machine learning framework utilizing manually gated training data that allows the integrative modeling of 'gated' cell types and the 'ungated' cells. By applying this framework on several CyTOF datasets, we demonstrated that including the 'ungated' cells can lead to a significant increase in the precision of the 'gated' cell types prediction. CyAnno can be used to identify even a single cell type, including rare cells, with higher efficacy than current state-of-the-art semi-automated approaches.

AVAILABILITY

The CyAnno is available as a python script with a user-manual and sample dataset at https://github.com/abbioinfo/CyAnno.

SUPPLEMENTARY INFORMATION

Supplementary data are available at Bioinformatics online.

Humoral and cellular immunogenicity and safety following a booster dose of a tetravalent dengue vaccine 5+ years after completion of the primary series in Singapore: 2-year follow-up of a randomized phase II, placebo-controlled trial

The tetravalent dengue vaccine (CYD-TDV) is approved for use as a 3-dose series for the prevention of dengue in seropositive individuals ≥9 years. A randomized, placebo-controlled, phase II study of a booster dose of CYD-TDV in individuals who completed the 3-dose schedule >5 years previously (NCT02824198), demonstrated that a booster restored neutralizing antibody titers to post-dose 3 levels. We present additional immunogenicity assessments up to 24 months post-booster, and B- and T-cell responses in a participant subset. Participants aged 9-45 years that had received all three doses of CYD-TDV were randomized 3:1 to receive a booster dose of CYD-TDV (n = 89) or placebo (n = 29). Neutralizing antibody levels at Months 1, 6, 12, and 24 post-booster were assessed by plaque reduction neutralization test. In a subset, B-cell responses were assessed by a fluorescent immunospot assay, and T-cells analyzed by flow cytometry at Days 0, 7, 12, Months 1 and 12. We observed an increase of antibody titers Month 1 post-booster, then a gradual decline to Month 24. In the CYD-TDV booster group, an increase in plasmablasts was seen at Day 7 declining by Day 14, an increase in memory B-cells was observed at Day 28 with no persistence at Month 12. CYD-TDV booster recalled a CD8+ T-cell response, dominated by IFN-γ secretion, which decreased 12 months post-booster. This study showed a short-term increase in antibody titers and then gradual decrease following CYD-TDV booster injection >5 years after primary immunization, and the presence of memory B-cells activated following the booster, but with low persistence.

2020 White Paper on Recent Issues in Bioanalysis: BAV Guidance, CLSI H62, Biotherapeutics Stability, Parallelism Testing, CyTOF and Regulatory Feedback (Part 2A - Recommendations on Biotherapeutics Stability, PK LBA Regulated Bioanalysis, Biomarkers Assays, Cytometry Validation & Innovation Part 2B - Regulatory Agencies' Inputs on Bioanalysis, Biomarkers, Immunogenicity, Gene & Cell Therapy and Vaccine)

The 14^th edition of the Workshop on Recent Issues in Bioanalysis (14^th WRIB) was held virtually on June 15-29, 2020 with an attendance of over 1000 representatives from pharmaceutical/biopharmaceutical companies, biotechnology companies, contract research organizations, and regulatory agencies worldwide. The 14^th WRIB included three Main Workshops, seven Specialized Workshops that together spanned 11 days in order to allow exhaustive and thorough coverage of all major issues in bioanalysis, biomarkers, immunogenicity, gene therapy and vaccine. Moreover, a comprehensive vaccine assays track; an enhanced cytometry track and updated Industry/Regulators consensus on BMV of biotherapeutics by LCMS were special features in 2020. As in previous years, this year's WRIB continued to gather a wide diversity of international industry opinion leaders and regulatory authority experts working on both small and large molecules to facilitate sharing and discussions focused on improving quality, increasing regulatory compliance and achieving scientific excellence on bioanalytical issues. This 2020 White Paper encompasses recommendations emerging from the extensive discussions held during the workshop, and is aimed to provide the Global Bioanalytical Community with key information and practical solutions on topics and issues addressed, in an effort to enable advances in scientific excellence, improved quality and better regulatory compliance. Due to its length, the 2020 edition of this comprehensive White Paper has been divided into three parts for editorial reasons. This publication covers the recommendations on (Part 2A) BAV, PK LBA, Flow Cytometry Validation and Cytometry Innovation and (Part 2B) Regulatory Input. Part 1 (Innovation in Small Molecules, Hybrid LBA/LCMS & Regulated Bioanalysis), Part 3 (Vaccine, Gene/Cell Therapy, NAb Harmonization and Immunogenicity) are published in volume 13 of Bioanalysis, issues 4, and 6 (2021), respectively.

An Overview of Flow Cytometry: Its Principles and Applications in Allergic Disease Research

Flow cytometry is a popular technique used for both clinical and research purposes. It involves laser-based technology to characterize cells based on size, shape, and complexity. Additionally, flow cytometers are equipped with the ability to take fluorescence measurements at multiple wavelengths. This capability makes the flow cytometer a practical resource in the utilization of fluorescently conjugated antibodies, fluorescent proteins, DNA binding dyes, viability dyes, and ion indicator dyes. As the technology advances, the number of parameters a flow cytometer can measure has increased tremendously, and now some has the capacity to analyze 30-50 or more parameters on a single cell. Here, we describe the basic principles involved in the mechanics and procedures of flow cytometry along with an insight into applications of flow cytometry techniques for biomedical and allergic disease research.

Standardization procedure for flow cytometry data harmonization in prospective multicenter studies

One of the most challenging objective for clinical cytometry in prospective multicenter immunomonitoring trials is to compare frequencies, absolute numbers of leukocyte populations and further the mean fluorescence intensities of cell markers, especially when the data are generated from different instruments. Here, we describe an innovative standardization workflow to compare all data to carry out any large-scale, prospective multicentric flow cytometry analysis whatever the duration, the number or type of instruments required for the realization of such projects.

Machine Learning of Discriminative Gate Locations for Clinical Diagnosis

High-throughput single-cell cytometry technologies have significantly improved our understanding of cellular phenotypes to support translational research and the clinical diagnosis of hematological and immunological diseases. However, subjective and ad hoc manual gating analysis does not adequately handle the increasing volume and heterogeneity of cytometry data for optimal diagnosis. Prior work has shown that machine learning can be applied to classify cytometry samples effectively. However, many of the machine learning classification results are either difficult to interpret without using characteristics of cell populations to make the classification, or suboptimal due to the use of inaccurate cell population characteristics derived from gating boundaries. To date, little has been done to optimize both the gating boundaries and the diagnostic accuracy simultaneously. In this work, we describe a fully discriminative machine learning approach that can simultaneously learn feature representations (e.g., combinations of coordinates of gating boundaries) and classifier parameters for optimizing clinical diagnosis from cytometry measurements. The approach starts from an initial gating position and then refines the position of the gating boundaries by gradient descent until a set of globally-optimized gates across different samples are achieved. The learning procedure is constrained by regularization terms encoding domain knowledge that encourage the algorithm to seek interpretable results. We evaluate the proposed approach using both simulated and real data, producing classification results on par with those generated via human expertise, in terms of both the positions of the gating boundaries and the diagnostic accuracy. © 2019 The Authors. Cytometry Part A published by Wiley Periodicals, Inc. on behalf of International Society for Advancement of Cytometry.

Immunobiotic and Paraprobiotic Potential Effect of Lactobacillus casei in a Systemic Toxoplasmosis Murine Model

One of the main characteristics of probiotics is their ability to stimulate and modulate the immune response regardless of their viability. Lactobacillus casei (Lc) can stimulate local and systemic immunity, in addition to the activation of macrophages at sites distant from the intestine. Activated macrophages limit the replication of intracellular protozoa, such as Toxoplasma gondii, through the production of nitric oxide. The present study aimed to evaluate the protection generated by treatment with viable and non-viable Lc in the murine systemic toxoplasmosis model. CD1 male mice were treated with viable Lc (immunobiotic) and non-viable Lc (paraprobiotic), infected with tachyzoites of Toxoplasma gondii RH strain. The reduction of the parasitic load, activation of peritoneal macrophages, inflammatory cytokines, and cell populations was evaluated at 7 days post-infection, in addition to the survival. The immunobiotic and paraprobiotic reduced the parasitic load, but only the immunobiotic increased the activation of peritoneal macrophages, and the production of interferon-gamma (IFN-γ), tumor necrosis factor (TNF), and interleukin-6 (IL-6) while the paraprobiotic increased the production of monocyte chemoattractant protein-1 (MCP-1) and T CD4⁺CD44⁺ lymphocytes. Viable and non-viable Lc increases survival but does not prevent the death of animals. The results provide evidence about the remote immunological stimulation of viable and non-viable Lc in an in vivo parasitic model.