Cyclone: an accessible pipeline to analyze, evaluate, and optimize multiparametric cytometry data.

Area of Science:

• Immunology
• Computational Biology
• Bioinformatics

Background:

• High-dimensional single-cell technologies are crucial for immunology research.
• Analyzing cytometry data requires computationally intensive clustering and dimensionality reduction.
• Current methods for evaluating and optimizing these analyses are challenging.

Purpose of the Study:

• To introduce Cytometry Clustering Optimization and Evaluation (Cyclone), an integrated analysis pipeline.
• To provide tools for dimensionality reduction, clustering, evaluation, and optimization of cytometry data.
• To facilitate the analysis and visualization of diverse cytometry datasets.

Main Methods:

• Developed the Cyclone analysis pipeline.
• Integrated dimensionality reduction, clustering, evaluation, and optimization tools.
• Applied Cyclone to mass cytometry (CyTOF), full-spectrum fluorescence cytometry, and multiplexed immunofluorescence (IF) data.

Main Results:

• Cyclone successfully recapitulated gold standard immune cell identification across multiple cytometry platforms.
• The pipeline enabled unsupervised identification of lymphocyte and mononuclear phagocyte subsets.
• Identified subsets were associated with distinct biological features in various contexts, including cancer and infectious diseases.

Conclusions:

• Cyclone is a versatile and accessible pipeline for cytometry data analysis.
• It optimizes and evaluates clustering for diverse datasets.
• The pipeline supports immunology research and biological discovery.