RMAExpress

RMAExpress computes gene expression summary values for Affymetrix Genechip® data using the Robust Multiarray Average (RMA) methodology to produce background-adjusted, quantile-normalized, and summarized probe-level expression measures for differential expression and quality assessment.

Key Features:

• Robust Multiarray Average (RMA): Implements background adjustment, quantile normalization, and summarization (median polish) to convert probe-level intensities into expression summaries.
• Probe-level normalization: Normalizes at the probe intensity level using complete-data methods that utilize information from all arrays to form a normalizing relation.
• Probe-level quality assessment: Computes probe-level metrics to evaluate expression measures and support detection of differentially expressed genes.
• Log transformation and model-based summarization: Applies log transformation to stabilize variance and supports model-based summarization such as the Li and Wong multiplicative model-based expression index (MBEI).
• Empirical validation: Methodologies validated with spike-in and dilution experiments demonstrating reduced bias and variance and improved detection compared to Affymetrix average difference (AvDiff) and MAS 5.0 signal.

Scientific Applications:

• Differential Gene Expression Analysis: Produces more accurate summary values to enhance detection of differentially expressed genes for studies of disease mechanisms, biomarker identification, and therapeutic target exploration.
• Quality Control in High-Density Oligonucleotide Arrays: Uses probe-level normalization and quality metrics to ensure reliable data for downstream analyses of high-density oligonucleotide (Affymetrix Genechip®) arrays.

Methodology:

Background adjustment, quantile normalization at the probe intensity level using complete-data methods across arrays, summarization via median polish, log transformation, model-based summarization (e.g., Li and Wong MBEI), and computation of probe-level quality metrics.