quantro

quantro evaluates the assumptions underlying quantile normalization to determine whether global distributional differences across sample groups reflect technical variability or biological signal.

Key Features:

• Assumption Testing: Tests whether observed global changes across samples are attributable to technical variability, the central assumption of quantile normalization.
• Data-Driven Approach: Provides an objective, data-driven assessment for the appropriateness of applying quantile normalization without relying on expert judgment.
• Group-Level Analysis: Uses user-supplied group labels (e.g., Tumor/Normal) to assess significant global distribution differences between groups and protect biological variation from being removed.
• Integration with Bioconductor: Implemented within Bioconductor to integrate with Bioconductor data structures and analysis workflows.

Scientific Applications:

• Gene expression analysis: Assesses whether quantile normalization is appropriate for high-throughput gene expression studies to help preserve biologically meaningful differential expression signals.
• DNA methylation analysis: Evaluates normalization assumptions for methylation profiles (e.g., MethylSet) to avoid removing biologically relevant methylation differences.
• Comparative studies across conditions: Guides normalization decisions in studies comparing conditions or treatments to ensure global distribution differences are not inadvertently normalized away.

Methodology:

Analyzes raw data objects (e.g., ExpressionSet, MethylSet) using group-level sample information to evaluate distributional differences between samples and test whether global changes are due to technical variability versus biological variation, with examples and simulations illustrating performance.