CellProfiler Analyst

CellProfiler Analyst analyzes quantitative image-derived data from high-throughput microscopy experiments to enable machine-learning-based phenotype classification and automated scoring.

Key Features:

• Supervised machine learning classifiers: Includes a supervised machine learning system trained on labeled cell/object examples to recognize intricate and subtle phenotypes.
• Neural network support: Adds support for neural network classifiers (introduced in version 3.0) enabling deep learning–based phenotypic classification and detection of rare object subsets.
• Automated high-throughput scoring: Automates scoring of millions of cells to scale phenotype quantification across large image-based experiments.
• Detection and measurement of object classes: Detects and measures specific classes of objects in images to generate quantitative features for downstream analysis.
• Interoperability with CellProfiler 4: Enables transfer of detected objects and measurements from CellProfiler 4 visualizations into the Classifier tool for use as training data.

Scientific Applications:

• Drug discovery: Supports high-throughput phenotypic profiling and compound screening by classifying cellular responses at scale.
• Phenotypic screening: Enables identification and scoring of diverse and subtle cellular phenotypes in large imaging screens.
• Cellular morphology and behavior research: Quantifies morphological and behavioral features of cells for basic biological investigations.

Methodology:

Integrates quantitative image analysis with supervised machine learning by training classifiers on labeled data, with additional support for neural network classifiers to enhance phenotype recognition and scoring.