sweetD

sweetD computes Hoeffding's D statistic to quantify and visualize the dependence between gene abundance (M) and log fold differences in expression (A) across large-scale gene expression datasets.

Key Features:

• Hoeffding's D Statistic: Quantifies non-parametric dependence between gene abundance (M) and log fold differences in expression (A) using Hoeffding's D.
• MA Plot Analysis: Uses MA plots (M versus A) to represent and assess relationships between gene intensity (abundance) and differential expression (log fold change).
• Scalability: Addresses the quadratic growth of MA plots with increasing sample numbers to enable analysis of large datasets.
• Visualization: Produces visual representations of M–A dependence to reveal patterns or anomalies such as batch effects and outliers.

Scientific Applications:

• Quality Control: Evaluates relationships between abundance and log fold differences to detect batch effects, outliers, or other sample-quality issues.
• Large-Scale Data Analysis: Applies to studies with many gene expression samples, including data generated by high-throughput sequencing technologies.

Methodology:

Calculates Hoeffding's D statistic across MA plots (M versus A) to quantify non-parametric dependence between gene abundance and log fold differences in expression.