Biases in Illumina transcriptome sequencing caused by random hexamer priming

Area of Science:

• Molecular Biology
• Genomics
• Bioinformatics

Background:

• cDNA generation using random hexamer priming is a common step in transcriptome sequencing.
• This method can introduce biases in nucleotide composition at the start of sequencing reads.
• Such biases affect the uniformity of read distribution across the transcriptome.

Purpose of the Study:

• To identify and characterize nucleotide biases in transcriptome sequencing reads generated by random hexamer priming.
• To develop a method to mitigate the impact of these biases on downstream analysis.

Main Methods:

• Analysis of nucleotide frequencies at the beginning of sequencing reads from Illumina Genome Analyzer.
• Development of a read count reweighting scheme based on observed nucleotide frequencies.
• Evaluation of the reweighting scheme's effectiveness in correcting bias and improving uniformity.

Main Results:

• Nucleotide composition bias at the start of reads is consistently observed, independent of organism or laboratory.
• The developed reweighting scheme effectively mitigates the impact of this bias.
• Improved uniformity of transcriptome reads was achieved after applying the reweighting method.

Conclusions:

• Random hexamer priming introduces a significant and consistent bias in transcriptome sequencing data.
• A nucleotide frequency-based reweighting approach can successfully correct for this bias.
• This method enhances the reliability and uniformity of transcriptome-wide read counts.