Related Experiment Video
Updated: May 16, 2026
05:22
Analyzing Multifactorial RNA-Seq Experiments with DiCoExpress
Published on: July 29, 2022
Analyzing illumina gene expression microarray data from different tissues: methodological aspects of data analysis in
Claudia Schurmann1, Katharina Heim, Arne Schillert
1Interfaculty Institute for Genetics and Functional Genomics, Ernst-Moritz-Arndt-University Greifswald, Greifswald, Germany.
Plos One
|December 14, 2012
Summary
Technical variations in gene expression data hinder meta-analysis. This study found that adjusting for key technical factors like Illumina chip assignment and RNA batch significantly improves the reliability of gene expression analyses, especially for large-scale studies.
Area of Science:
- Molecular Biology
- Functional Genomics
- Bioinformatics
Background:
- Microarray gene expression profiling is crucial but challenged by experimental and technical variability.
- Meta-analysis of gene expression studies requires robust data preprocessing and understanding of variation sources.
Purpose of the Study:
- To investigate the impact of data preprocessing and sample processing on gene expression data variability.
- To identify key technical and biological factors influencing gene expression signals in population-based cohorts.
- To assess the effect of single nucleotide polymorphisms (SNPs) on probe-specific expression signals.
Main Methods:
- Analysis of 3358 samples from German population-based cohorts using Illumina HumanHT-12 v3 BeadChip arrays.
- Comparison of log(2) and variance-stabilizing transformations for gene expression data preprocessing.
- Principal Component Analysis (PCA) to identify sources of variation.
- Assessment of technical factors (chip assignment, RNA amplification/isolation batches, storage time) and biological factors (cell composition, sex).
- Evaluation of SNP effects on probe hybridization and signal intensity.
Main Results:
- Log(2) and variance-stabilizing transformations yielded similar gene expression signal intensities.
- Technical factors, particularly Illumina chip assignment (33-48% variance) and RNA amplification batch (12-24%), were major contributors to data variability.
- Sample processing time (blood to RNA, RNA to amplification) also influenced signals.
- White blood cell composition (3%) and sex (1-2%) were the strongest biological influences.
- Most SNPs within probes (38% of probes) had minimal impact on hybridization efficiency (1.4% showed significant modification).
Conclusions:
- Adjustment for specific technical factors significantly enhances the reliability of gene expression data.
- These adjustments are essential for accurate meta-analyses of gene expression studies.
- Understanding and mitigating technical variability is key to advancing molecular biology and functional genomics research.
