Researchers at UC Berkeley have shown that priming using random hexamers biases the nucleotide content of RNA-Seq reads and that this bias also affects the uniformity of the locations of the reads along expressed transcripts.
Despite this bias, they suggest that random hexamer priming is still preferable to the alternative oligo(dT) priming (which is highly biased toward the 3′-end of the expressed transcripts) because the impact of hexamer priming can be mitigated. They provide a read count reweighting scheme, based on the nucleotide frequencies of the reads, to achieve this.
- Hansen KD, Brenner SE, Dudoit S. (2011) Biases in Illumina transcriptome sequencing caused by random hexamer priming. Nucleic Acids Res 38(12), e131. [article]