flowUtils

flowUtils performs model-based clustering of flow cytometry (FCM) data to identify and extract cell populations using multivariate t mixture models with Box-Cox transformation and explicit outlier detection to handle data variability and high dimensionality.

Key Features:

• Automated Analysis: Automates population identification to reduce reliance on manual sequential gating in 2-D projections.
• Model-Based Clustering: Implements multivariate t mixture models for robust clustering of FCM measurements.
• Box-Cox Transformation: Applies Box-Cox transformation to stabilize variance and improve clustering accuracy.
• Outlier Identification: Detects and manages outliers to improve the quality of extracted cell populations.
• Data Transformation Handling: Provides explicit handling of common data transformation challenges encountered in FCM analysis.
• Visualization of Clustering Results: Produces visual summaries to explore and interpret clustering outcomes.

Scientific Applications:

• Cell population identification: Extraction and characterization of distinct cell populations from complex FCM datasets.
• Health research: Analysis of high-throughput single-cell measurements for biomedical studies.
• Medical diagnosis and treatment: Support for diagnostic immunophenotyping and monitoring of treatment responses.
• Vaccine development: High-dimensional cytometry analyses to evaluate immune responses in vaccine research.

Methodology:

Model-based clustering using multivariate t mixture models combined with Box-Cox transformation and outlier detection to automate population identification in high-dimensional flow cytometry data, replacing manual sequential gating.