diffcyt: Differential discovery in high-dimensional cytometry via high-resolution clustering.

Area of Science:

• Immunology
• Computational Biology
• Bioinformatics

Background:

• High-dimensional cytometry, including mass and flow cytometry, enables detailed single-cell analysis of over 40 protein markers.
• Analyzing large, high-dimensional cytometry datasets is challenging due to data complexity and limitations of current computational tools.

Purpose of the Study:

• To introduce diffcyt, a novel computational framework designed for differential discovery analyses in high-dimensional cytometry data.
• To address the analytical challenges posed by the scale and dimensionality of cytometry datasets.

Main Methods:

• diffcyt combines high-resolution clustering with empirical Bayes moderated tests, adapting methods from transcriptomics.
• The framework is implemented in an open-source environment for accessibility and broad application.

Main Results:

• The diffcyt approach demonstrates enhanced statistical performance for differential analyses.
• Improved detection of rare cell populations is a key advantage of this new method.
• The framework supports flexible experimental designs and offers fast computational runtimes.

Conclusions:

• diffcyt provides a robust and efficient computational solution for high-dimensional cytometry data analysis.
• This framework enhances the ability to discover and characterize cell populations, particularly rare ones.
• The open-source nature of diffcyt promotes wider adoption and advancement in single-cell cytometry research.