The T7-primer is a source of experimental bias and introduces variability between microarray platforms

Area of Science:

• Molecular Biology
• Genomics
• Bioinformatics

Background:

• Eberwine-like amplification is a common method for amplifying RNA.
• This process involves adding nucleotide sequences to RNA transcripts.
• Potential biases introduced during amplification can affect downstream analyses.

Purpose of the Study:

• To investigate the impact of nucleotide stretches added during Eberwine-like amplification on microarray signals.
• To identify the cause of aberrantly high signals observed in specific probes.
• To assess the implications of T7 primer bias for cross-platform data comparability.

Main Methods:

• Analysis of over six thousand microarray datasets.
• Comparison of signals from probes targeting specific nucleotide motifs against unaffected probes.
• Evaluation of amplification products derived from different T7 primer sequences.

Main Results:

• Eberwine-like amplification introduces 6-10 bp nucleotide stretches at the 5' end of RNA transcripts.
• Probes complementary to these stretches show up to 100-fold higher signals compared to unaffected probes.
• This artifact is specific to amplified RNA and not observed with total RNA.
• Significant T7 primer bias exists across different datasets and platforms.

Conclusions:

• The nucleotide stretches added during Eberwine-like amplification create a significant bias in microarray analysis.
• This bias can lead to overestimation of gene expression, particularly for probes targeting these added sequences.
• The observed T7 primer bias poses a challenge for comparing microarray data across different laboratories and platforms, necessitating careful data normalization and validation strategies.