GatingTree: Navigating Multidimensional CytometryData

Dr. Masahiro Ono

2024-11-04

GatingTree is an R package designed to enhance the analysis of group-specific effects in cytometry data.

Current Challenges

• Advancements in Technology: Cytometry technologies have advanced to allow simultaneous analysis of numerous markers.
• Complex Data Analysis: These advancements introduce challenges in analyzing biological effects, such as treatment effects, within high-dimensional data.
• Limitations of Traditional Methods:
  • Manual gating does not meet complex analytical demands.
  • Dimensional reduction methods like UMAP, with or without clustering, are commonly used but struggle with:
    • Reproducibility: Poor reproducibility across experiments.
    • Practical Application: Identified cell clusters often do not translate into effective gating strategies in the lab.
• Reproducibility Crisis: These issues contribute to a reproducibility crisis in biological and medical research.

Solutions GatingTree Offers

• Innovative Approach:
  • No Dimensional Reduction: Does not rely on dimensional reduction techniques.
  • Pathfinding Analysis: Explores the multidimensional marker space to pinpoint group-specific features.
• Avoidance of Standard Analyses:
  • Deliberately avoids traditional multidimensional analyses like PCA and UMAP.
  • Does not use clustering algorithms.
• Practical Applications: Provides straightforward solutions that are:
  • Directly applicable in downstream processes.
  • Ideal for tasks such as flow cytometric sorting of target populations.