flowAI

flowAI performs automated quality control of flow cytometry (FCM) data as an R package, detecting and removing anomalies in flow rate, signal acquisition, and dynamic-range outliers to improve downstream analyses.

Key Features:

• R package: Implemented as an R package for processing flow cytometry data files.
• Automatic anomaly detection: Implements algorithms to detect and remove unwanted events from FCM files.
• Interactive/manual refinement: Provides an interactive mode for manual refinement of data quality.
• Flow rate analysis: Identifies abrupt changes in the flow rate that indicate anomalous segments.
• Signal acquisition stability: Evaluates stability of signal acquisition to flag inconsistent data capture.
• Dynamic range assessment: Detects outliers at the lower limit and margin events at the upper limit of the dynamic range.
• Quality assessment summary: Generates a detailed per-file summary report of the quality assessment results.

Scientific Applications:

• Cell phenotype and function characterization: Cleans FCM data to support accurate identification of cell phenotypes and functions.
• Unbiased cell sub-population identification: Improves reliability of clustering and dimensionality reduction for segregation of cell sub-populations.
• Clinical and basic research: Supports analyses relevant to clinical diagnostics and basic biological research using high-dimensional FCM datasets.

Methodology:

Computational steps include flow rate analysis to detect abrupt changes, evaluation of signal acquisition stability, and dynamic-range outlier detection to identify and remove anomalous events.