On existing short-read sequencing platforms, the main bottleneck in processing large numbers of samples is preparing them for loading: there is an acute need for low-cost, high-throughput, highly-multiplexed library production methods that moreover require only small amounts of input material. Such high outputs are essential for human whole-genome sequencing but excessive for many other applications, where the same amount of data would be sufficient to sequence several samples in applications such as targeted sequencing, microbial genome sequencing, RNAseq, ChIP-seq and amplicons. In the race for the first $1,000 human genome, next-generation, high-throughput sequencers such as the Illumina HiSeq instrument have been developed that can produce tens of gigabases of raw sequence per day.

We addressed the bottlenecks of Nextera library construction to produce a modified protocol which harnesses the full power of the Nextera kit and allows the reproducible construction of libraries on a high-throughput scale reducing the associated cost of the kit. Finally we developed a custom set of index tags and primers which increase the number of samples that can simultaneously be sequenced on a single lane of an Illumina instrument.

A 75% cost reduction was achieved with a low-volume modified protocol which was tested over genomes with different GC content to demonstrate its robustness. We show that an initial bead-based normalisation step can remove the need for quantification and improves sample purity. We addressed the identified limitations to develop a robust workflow that supports both rapid and high-throughput projects also reducing reagent costs.

The Nextera protocol, which utilises a transposome based approach to create libraries for Illumina sequencing, requires pure DNA template, an accurate assessment of input concentration and a column clean-up that limits its applicability for high-throughput sample preparation.

Download two dots level 104