Large Scale Library Generation for High Throughput Sequencing Authors and Affiliations

Background: Large efforts have recently been made to automatethe sample preparation protocols for massively parallel sequencing in order to match the increasing instrument throughput. Still, the size selection through agarose gel electrophoresis separation is a labor-intensive bottleneck of these protocols. Methodology/Principal Findings: In this study a method for automatic library preparation and size selection on a liquid handling robot is presented. The method utilizes selective precipitation of certain sizes of DNA molecules on to paramagnetic beads for cleanup and selection after standard enzymatic reactions. Conclusions/Significance: The method is used to generate libraries for de novo and re-sequencing on the Illumina HiSeq 2000 instrument with a throughput of 12 samples per instrument in approximately 4 hours. The resulting output data show quality scores and pass filter rates comparable to manually prepared samples. The sample size distribution can be adjusted for each application, and are suitable for all high throughput DNA processing protocols seeking to control size intervals.