High-throughput single cell data analysis - A tutorial.

Area of Science:

• Immunology
• Biotechnology
• Environmental Science

Background:

• White blood cells (WBCs) are vital for immunity but implicated in diseases like leukemia and autoimmune disorders.
• Measuring protein expression on millions of single cells via high-throughput instruments is key to understanding WBC function.
• Extracting comprehensive biochemical information necessitates advanced multivariate statistical methods.

Purpose of the Study:

• To provide an overview of essential multivariate data analysis steps for single-cell data.
• To demonstrate these methods using white blood cells (immunology) as a case study.
• To highlight the applicability of these methods in environmental and biotech research.

Main Methods:

• Data analysis begins with study design assessment and research question formulation.
• Key designs include immunophenotyping, cell activation studies, and rare cell discovery.
• Data pre-processing involves focusing on the cell type of interest and converting single-cell data into cellular distributions.

Main Results:

• Clustering methods like Self-Organizing Maps are suitable for immunophenotyping.
• Principal Component Analysis is effective for modeling covariance in cell activation studies.
• Rare cell discovery involves modeling common cells and then identifying rare populations.

Conclusions:

• Discriminant analysis of cellular distributions can identify differential cell subtypes between groups.
• The presented multivariate approach is broadly applicable to single-cell data analysis across various research fields.
• This framework enhances the extraction of meaningful biological insights from complex high-throughput datasets.