Related Experiment Video
Updated: Jun 5, 2025
05:12
ExCYT: A Graphical User Interface for Streamlining Analysis of High-Dimensional Cytometry Data
Published on: January 16, 2019
11.4K
Comprehensive evaluation and practical guideline of gating methods for high-dimensional cytometry data: manual
Peng Liu1, Yuchen Pan2, Hung-Ching Chang1
1Department of Biostatistics, School of Public Health, University of Pittsburgh, 130 De Soto St., Pittsburgh, PA 15261, US.
Briefings in Bioinformatics
|December 10, 2024
Summary
This study compares cytometry data analysis methods, including manual gating, unsupervised clustering, and supervised auto-gating. Several unsupervised tools like PAC-MAN and supervised methods like DeepCyTOF showed strong performance for cell population identification.
Area of Science:
- Biotechnology
- Computational Biology
- Immunology
Background:
- Cytometry enables single-cell resolution protein identification and quantification.
- High-dimensional cytometry data analysis requires accurate cell population phenotyping based on marker expressions.
Purpose of the Study:
- To quantitatively review and compare manual gating, unsupervised clustering, and supervised auto-gating methods for cytometry data.
- To provide recommendations for selecting appropriate phenotyping methods based on application scenarios.
Main Methods:
- Evaluation of manual gating consistency across researchers.
- Quantitative comparison of 23 unsupervised clustering tools for accuracy and computational cost.
- Assessment of four supervised auto-gating algorithms.
Main Results:
- Manual gating showed inter-rater variability.
- Unsupervised tools PAC-MAN, CCAST, FlowSOM, flowClust, and DEPECHE demonstrated strong performance.
- Supervised methods DeepCyTOF and CyTOF Linear Classifier achieved the best results.
Conclusions:
- No single method universally outperformed others.
- Specific unsupervised and supervised methods offer robust solutions for cytometry data analysis.
- Guidance is provided for biologists to choose optimal gating strategies for their research.
