TGx-DDI

TGx-DDI classifies DNA damage responses from transcriptomic data to support genotoxicity hazard assessment of pharmaceuticals and environmental chemicals.

Key Features:

• Machine Learning-Driven Identification: Uses machine learning to identify a specific set of transcripts responsive to DNA damage, originally described as TGx-28.65.
• Integration with Testing Paradigms: Applies alongside existing genotoxicity testing batteries to help distinguish clastogenic agents that are harmful in vivo from those yielding positive results in in vitro chromosomal damage (CD) assays.
• High-Throughput Implementation: Implemented in a high-throughput cell-based genotoxicity testing system utilizing NanoString nCounter® technology for large-sample processing.
• Validation and Reproducibility: Performance has been assessed for intra- and inter-laboratory reproducibility and cross-platform consistency.
• Regulatory Qualification Efforts: Subject to qualification efforts at the US Food and Drug Administration (FDA) for use in drug development genotoxicity testing strategies.

Scientific Applications:

• Discrimination of Genotoxicants: Distinguishes true DNA-damaging agents from irrelevant positives in in vitro chromosomal damage (CD) assays.
• Hazard Assessment: Supports genotoxicity hazard assessment of pharmaceuticals and environmental chemicals using transcriptomic signatures.
• High-Throughput Screening: Enables large-scale transcriptomic genotoxicity screening workflows via NanoString nCounter® implementation.

Methodology:

Machine learning was applied to transcriptomic data to identify the TGx-28.65 transcript set, and performance was evaluated through intra- and inter-laboratory and cross-platform analyses.