Removing unwanted variation with CytofRUV to integrate multiple CyTOF datasets

Area of Science:

• Single-cell biology
• Immunology
• Computational biology

Background:

• Mass cytometry (CyTOF) is a powerful technology for high-dimensional single-cell analysis.
• Characterizing cell subpopulations requires analyzing millions of single cells.
• Integrating data from multiple CyTOF batches acquired across different sites and times presents a significant challenge due to technical variability.

Purpose of the Study:

• To develop and present a robust computational approach for integrating mass cytometry datasets from multiple batches.
• To address the technical limitations hindering the comparison of CyTOF data across different experimental runs.
• To enable confident comparison of cellular changes and correlation with clinical outcomes.

Main Methods:

• Development of CytofRUV, a novel approach for analyzing and correcting batch effects in mass cytometry data.
• Implementation of an R-Shiny application with diagnostic plots to visualize and assess batch correction.
• Application of CytofRUV to integrate large-scale CyTOF datasets from multiple patients and conditions.

Main Results:

• CytofRUV effectively corrects for batch effects in mass cytometry data.
• The approach facilitates the integration of data from numerous patients and conditions across different batches.
• Enables confident comparison of cellular subpopulations and their alterations.

Conclusions:

• CytofRUV provides a reliable solution for integrating multi-batch mass cytometry data.
• This method overcomes a major hurdle in single-cell biology, enabling more comprehensive and accurate analyses.
• Facilitates robust correlation of cellular phenotypes with clinical outcomes.