CytoPipeline and CytoPipelineGUI: a Bioconductor R package suite for building and visualizing automated...

Area of Science:

• Bioinformatics
• Computational Biology
• Data Science

Background:

• High-dimensional flow cytometry data necessitates automated analysis pipelines to complement manual gating.
• Effective pre-processing, including compensation, scaling, and cell quality filtering, is critical for downstream analysis accuracy.
• Challenges in pre-processing include method variability, step order impact, and non-standardized quality control visualizations.

Purpose of the Study:

• To introduce CytoPipeline and CytoPipelineGUI, R packages designed for building, comparing, and assessing flow cytometry pre-processing pipelines.
• To provide a framework for users to design and evaluate pre-processing workflows on real-world datasets.
• To enhance the user's ability to gain insights from benchmarking pre-processing pipelines.

Main Methods:

• Development of two Bioconductor R packages: CytoPipeline for pipeline construction and CytoPipelineGUI for graphical user interface.
• Demonstration of pipeline design and visual assessment using a real-life flow cytometry dataset.
• Benchmarking of two distinct pre-processing pipelines with varying quality control methods.

Main Results:

• CytoPipeline and CytoPipelineGUI facilitate the creation and evaluation of automated pre-processing workflows.
• Visual assessment tools within the packages offer intuitive insights into pipeline performance.
• Benchmarking results highlight the impact of different quality control strategies on data analysis.

Conclusions:

• CytoPipeline and CytoPipelineGUI streamline the development and testing of flow cytometry pre-processing pipelines.
• These packages improve user productivity by offering intuitive visualizations for assessing pipeline quality and benchmarking results.
• The tools complement existing benchmarking approaches by providing clear, user-friendly insights into complex data pre-processing steps.