Jones, M.B. et al. Proc. Natl. Acad. Sci. USA 112, 14024–14029 (2015).

As shotgun metagenomic sequencing becomes more widespread, there is an increasing need to understand biases across different technologies and to determine whether studies based on different platforms are comparable. Jones et al. used a synthetic mixture, or mock community, of microbial sequences to benchmark library preparation using PCR-free Illumina Nextera XT and TruSeq DNA kits as well as the KAPA Biosystems Hyper Prep PCR and PCR-free kits. The authors found that a high GC base content can inflate biases in error profile, duplication rate and read loss, and they recommend using PCR-free kits to reduce bias and cell spike-in controls for more accurate quantification. They also studied the effects of the different library-preparation methods on human stool samples.