ORIOGEN

ORIOGEN performs order-restricted inference on time-course and dose-response gene expression data to estimate mean expression profiles and identify genes with temporal or dose-dependent changes under heteroscedastic variability.

Key Features:

• Iterative algorithm: Employs an iterative algorithm to estimate mean gene expression levels across time points or dose levels guided by predefined order-restrictions.
• Order-restrictions: Uses predefined patterns or profiles (order-restrictions) to constrain estimation of temporal or dose-dependent dynamics.
• Handling heteroscedasticity: Accounts for heteroscedasticity in gene expression data to provide robust inference when variability changes over time or with dose.
• Bootstrap-based gene selection: Implements a bootstrap-based procedure for gene selection that controls the false positive rate at the nominal level.
• Comparative performance: Simulation studies demonstrate the enhanced methodology competes favorably with the original ORIOGEN in statistical power while maintaining stringent control over error rates.

Scientific Applications:

• Oncology and pharmacology studies: Applied to contexts requiring detection of dynamic gene expression responses to treatment over time or across dose levels.
• Breast cancer cell-line analysis: Used to identify genes responding to treatment in breast cancer cell-line datasets in time-course or dose-response experiments.

Methodology:

Uses an iterative algorithm for order-restricted estimation of mean expression across time or dose levels, accounts for heteroscedasticity, applies bootstrap-based gene selection controlling the nominal false positive rate, and was evaluated by simulation studies comparing statistical power and error control.