RTCA

RTCA classifies time-dependent cellular response profiles from Roche xCELLigence real-time cell analysis (RTCA) data to assess cytotoxicity and infer chemical modes of action (MoA).

Key Features:

• BioConductor package: Implemented as the 'rtca' package on BioConductor for analysis of RTCA datasets.
• Data input: Processes multi-concentration time-dependent cellular response profiles (TCRPs) captured by the xCELLigence RTCA HT system.
• Model-based hierarchical classification: Uses a hierarchical classification framework to group similar cytotoxic response patterns into classes.
• Dimensionality reduction (PCA): Applies principal component analysis (PCA) to identify key patterns and reduce dimensionality in complex datasets.
• Functional data analysis (FDA): Models temporal response curves as continuous functions to capture dynamics over time.
• Mode-of-action classification: Assigns cytotoxic agents to classes based on similarity of response patterns to infer chemical MoA.
• Transformation strategies: Provides a suite of data transformation methods for preparing RTCA time-series data for analysis.
• Visualization: Includes visualization tools tailored for RTCA time-course and classification results.
• Biological validation: Validation demonstrated by biological verification using human hepatocellular carcinoma cells (HepG2) and examination of common chemical mechanisms of action.

Scientific Applications:

• Cytotoxicity assessment: Quantifies and classifies cytotoxic responses measured by real-time cell analysis.
• Toxicity profiling of unclassified agents: Rapidly groups uncharacterized toxic agents by similarity to known MoAs for hypothesis generation.
• High-throughput screening analysis: Analyzes RTCA HT multi-concentration time-course data for large-scale toxicity screens.
• Mode-of-action discovery: Enables identification and discrimination of distinct chemical modes of action from dynamic response patterns.
• Time-series response modeling: Supports analysis of continuous temporal dynamics in cellular response profiles across concentrations.

Methodology:

Model-based hierarchical classification leveraging principal component analysis (PCA) and functional data analysis (FDA) to reduce dimensionality of multi-concentration TCRPs and model responses as continuous functions over time from xCELLigence RTCA HT data.