CDS-DB

CDS-DB curates patient-derived, drug-induced gene expression signatures from paired pre- and post-treatment clinical transcriptomes to enable analysis of treatment-associated molecular responses in cancer.

Key Features:

• Data collection and standardization: Collects 78 datasets of paired pre- and post-treatment transcriptomic profiles uniformly reprocessed with manually curated metadata that includes drug administration dosage, sampling time and location, and intrinsic drug response status.
• Extensive coverage: Comprises 2,012 patient-level gene perturbation signatures spanning 85 therapeutic regimens and 39 cancer subtypes derived from 3,628 patient samples.
• Single-signature analyses: Provides differential gene expression analysis, functional enrichment, tumor microenvironment assessment, and correlation analyses for individual signatures.
• Comparative and connectivity analyses: Supports comparative signature analysis and connectivity analysis to investigate drug mechanisms of action (MOAs) and inter-patient heterogeneity.

Scientific Applications:

• Drug mechanisms of action (MOA): Enables identification of molecular pathways and perturbation patterns underlying drug MOAs from treatment-induced expression changes.
• Understanding drug resistance: Facilitates investigation of molecular mechanisms associated with drug resistance or sensitivity in tumors using paired pre- and post-treatment signatures.
• Drug repositioning and discovery: Supports discovery of drug repositioning opportunities and novel drug combinations through signature comparison and connectivity analyses.

Methodology:

Datasets were uniformly reprocessed and analyses explicitly include differential gene expression, functional enrichment, tumor microenvironment assessment, correlation analysis, comparative signature analysis, and connectivity analysis.