Combining gene expression data from different generations of oligonucleotide arrays.

Area of Science:

• Genomics
• Bioinformatics
• Microarray Analysis

Background:

• Leveraging accumulated microarray data is crucial for comprehensive biological insights.
• Genomic annotation and probe design changes hinder data comparability across microarray platforms.

Purpose of the Study:

• To address the challenge of comparing gene expression data between different microarray generations.
• To develop and validate a method for enhancing data comparability in microarray analysis.

Main Methods:

• Comparative analysis of two Affymetrix oligonucleotide array generations (HG-U95Av2 and HG-U133A) using human muscle biopsy samples.
• Evaluation of probe matching strategies (Affymetrix table, UniGene, LocusLink) and data preprocessing techniques (rescaling, filtering).
• Development of a probe filtering method based on overlapping sequence segments between array types.

Main Results:

• Significant discordance observed in cluster analysis and differential gene expression identification between array generations.
• Standard probe matching and data preprocessing methods show limited success in improving comparability.
• Probe filtering based on sequence overlap significantly enhances data comparability while retaining sufficient probe sets.

Conclusions:

• Probe-level sequence information critically impacts high-density oligonucleotide array compatibility.
• A carefully filtered subset of probes based on sequence overlaps enables more effective integration of data from different microarray generations.