CITEViz: interactively classify cell populations in CITE-Seq via a flow cytometry-like gating workflow using R-Shiny.

Area of Science:

• Genomics
• Immunology
• Bioinformatics

Background:

• Multi-omic single-cell sequencing assays, like Cellular Indexing of Transcriptomes and Epitopes by Sequencing (CITE-Seq), integrate multiple data types from individual cells.
• CITE-Seq combines RNA transcriptome and surface protein expression profiling, offering deeper biological insights than single-modality approaches.
• Identifying cell populations through surface protein markers (gating) is crucial for CITE-Seq data analysis but can be complex and require extensive coding.

Purpose of the Study:

• To develop CITEViz, an R-Shiny application designed to simplify and standardize the cell gating process for CITE-Seq data.
• To provide an interactive platform for gating cells within Seurat-processed CITE-Seq datasets.
• To offer integrated visualization of quality control (QC) metrics for holistic CITE-Seq data evaluation.

Main Methods:

• CITEViz was developed as an R-Shiny application.
• The tool processes Seurat-object formatted CITE-Seq data.
• Interactive gating is performed using surface protein markers, analogous to flow cytometry gating.

Main Results:

• CITEViz was successfully applied to a peripheral blood mononuclear cell CITE-Seq dataset, enabling gating of major blood cell populations.
• The application facilitated the investigation of cellular heterogeneity within monocyte subsets and identified donor-specific antibody detection variations.
• Visualization tools within CITEViz aided in the robust classification of single-cell populations.

Conclusions:

• CITEViz standardizes the gating workflow for CITE-Seq data, enhancing the efficiency of cell population classification.
• The application generates essential feature plots and QC figures tailored for multi-omic data.
• CITEViz integrates user-friendly interface design with robust data structures, facilitating data retrieval and analysis for both biologists and computational scientists, ultimately aiding hypothesis generation.