flowMeans

flowMeans applies modified K-means clustering and change point detection to automatically identify and segment cell populations in flow cytometry (FCM) data, enabling detection of concave and complex population shapes.

Key Features:

• Non-Parametric Clustering: Uses a modified K-means algorithm that models a single cell population with multiple clusters to capture concave and complex population shapes.
• Change Point Detection: Integrates a change point detection algorithm to determine the optimal number of sub-populations within a dataset.
• Time Efficiency and Accuracy: Optimized for computational efficiency and reports accuracy comparable to manual gating and other automated gating algorithms.

Scientific Applications:

• High-throughput FCM studies: Supports large-scale flow cytometry datasets requiring efficient and automated population identification.
• Immunology and oncology: Enables precise identification of cellular subpopulations relevant to immunological and oncological research.
• Automated gating and reproducibility: Automates gating to reduce human error and bias, improving consistency and reproducibility across analyses.

Methodology:

Applies a two-step computational approach: modified K-means clustering to model complex-shaped cell populations, followed by change point detection to refine clusters into distinct sub-populations.