Identification of artifactual microarray probe signals constantly present in multiple sample types.

Area of Science:

• Molecular Biology
• Genomics
• Bioinformatics

Background:

• Transcript detection methods have advanced from Northern analysis to RNA sequencing (RNA-Seq).
• Microarrays and RNA-Seq are common tools for gene expression profiling.
• Discrepancies between microarray and RNA-Seq data can impact research findings.

Purpose of the Study:

• To identify and characterize discordant probes between microarray and RNA-Seq data.
• To assess the reliability of specific microarray probes for transcript quantitation.
• To investigate the array-dependency of probe performance.

Main Methods:

• Comparative analysis of microarray and RNA-Seq data across four human tissues/cell lines.
• Identification of probes with consistently high signal intensity discordant with RNA-Seq.
• Validation of transcript absence using quantitative RT-PCR (qRT-PCR).

Main Results:

• Ninety-nine out of ~30,000 microarray probes were identified as consistently discordant.
• Seventy of these discordant probes showed high signals across 713 samples on the Illumina HumanHT-12v4 platform.
• Some discordant probes were array-dependent and showed discrepancies with other probes for the same gene.
• qRT-PCR confirmed the absence of several targeted transcripts.

Conclusions:

• Certain microarray probes may provide misleading high signal intensities, not reflecting true transcript levels.
• Researchers should exercise caution when interpreting data from specific microarray probes.
• Validation with orthogonal methods like RNA-Seq or qRT-PCR is crucial for accurate transcript quantitation.