Video Poster: High-throughput HiFi library workflow for human whole genome sequencing on the Sequel II System
Recent advances in sequencing chemistry and software in the Sequel II System enable generating highly accurate long reads that are up to 25 kb in length with >99% accuracy. The high quality HiFi reads are suitable for variant detection of all types, from single nucleotides to structural variants. PacBio offers an end-to-end solution from sample preparation to data analysis. However, library construction is still a bottleneck making it difficult to implement into a high-throughput workflow for sequencing large number of samples. Input DNA requirements, DNA shearing and size-selection/fractionation are the most critical and challenging steps in the current procedure. In this study we will address each of these steps and highlight alternate methods that can be implemented to make the procedure scalable for any automated liquid handling systems. At present, the HiFi library preparation workflow requires 10-15 µg of human genomic DNA, which may be prohibitive for WGS projects where source of DNA is scarce (e.g., DNA extracted from infant blood). As a solution, we have optimized the workflow using 1-2 µg of human genomic DNA generating adequate data with >20-fold genome coverage per sample. DNA was fragmented to 15 kb – 20 kb using Diagenode’s Megaruptor 3 which can process up to 8 samples simultaneously. In addition, library construction using the SMRTbell Express Template Kit 2.0 ensured increased sample recovery with faster turn-around time. Above all, we eliminated the time-consuming size-selection process and employed AMPure PB beads for size-selection making it amenable to automation. The resulting HiFi SMRTbell library generated >24 GB HiFi reads per Sequel II SMRT Cell 8M. The mean CCS read lengths were ~20 kb with read quality of Q31 (median).