flowEMMi

flowEMMi performs automated model-based clustering of microbial flow cytometry (FCM) data to identify and separate cell subpopulations within complex microbial communities.

Key Features:

• Model-based clustering: Uses multivariate Gaussian mixture models for clustering of FCM data.
• Subsampling: Employs subsampling to reduce computational load during cluster estimation.
• Foreground/background separation: Implements foreground/background separation to distinguish cells from technical noise, beads, and cell debris.
• Two-channel, high-throughput support: Designed for high-throughput, two-channel flow cytometry datasets.
• Runtime heuristics: Includes optional heuristics to decrease running times and enable near-online results.
• Performance: Provides fast and accurate cluster identification compared with other available methods.
• Information-rich outputs: Produces clustering results with high information content and consistency validated by ancillary data and statistical proof.
• Automation: Automates cluster identification to reduce manual clustering and user input.

Scientific Applications:

• Microbial community analysis: Identifying and characterizing cell subpopulations within complex microbial communities using FCM.
• Subcommunity dynamics: Tracking temporal dynamics of microbial subcommunities in environmental or experimental time series.
• Monitoring and near-online analysis: Enabling near-online monitoring of microbial community changes in high-throughput experiments.
• Health-related samples: Applying microbial FCM analysis in contexts related to medical diagnostics and health research.

Methodology:

Performs model-based clustering using multivariate Gaussian mixture models with subsampling, foreground/background separation, and optional runtime heuristics for two-channel FCM data.