Flexible, production-scale, human whole genome sequencing on a benchtop sequencer

Abstract

Human whole-genome sequencing (hWGS) provides comprehensive genomic information that can help guide research in disease prevention and treatment. Recent advancements in sequencing technology have improved sequencing quality and further reduced sequencing costs on bench-top sized instruments, making whole-genome sequencing an accessible technology for broader use. Here, we demonstrate the feasibility of a large WGS project using a benchtop sequencer in a small laboratory setting, on a scale previously reserved for production-scale machines. In this project, 807 samples were prepared and sequenced across 313 flow cells, with high sequencing quality at a median %Q30 of 96.6% and a median %Q40 of 89.31%. To screen library quality and maximize sample yield, we utilized 48-plex sample pre-pool 'QC' runs to provide > 1 × coverage per sample prior to sample pooling and full-depth sequencing, providing valuable sample-level insights prior to full-depth sequencing. With this strategy, we consistently achieved > 30 × human whole genome sequencing of three-plex sample trios with standard settings. To demonstrate additional flexibility present in the platform, we explored two different use cases 1) large insert sizes (1kb +) library to achieve superior genome coverage; 2) proof of concept rapid WGS sequencing to minimize sample to answer turnaround time for time-critical sequencing applications. Sequencing of a 2 × 100 > 30 × human WGS can be achieved in < 12 h and subsequent file generation in < 1 additional hour. This study provides a cost-effective and flexible real-world demonstration of achieving both high quality hWGS sequencing and instrument flexibility without the need for complex batching schemes or factory-sized sequencers.

Keywords: Bench Top Sequencer; Benchmarking; Genomics; High-Throughput Sequencing; Human Whole Genome Sequencing; Rapid Whole Genome Sequencing; Sequencing.

Fig. 1

Laboratory workflow from sample ingest to data analysis

Fig. 2

A Median aligned insert size of the small (386 bp, SD 122.64 bp), medium (584 bp, SD 170.10 bp), and large (1339 bp, SD 444.26 bp) insert libraries. Insert size is calculated by alignment to HG001 reference genome. B Base quality remained high across all cycles of the run. C Mean base quality by run for each insert size

Fig. 3

Increase in library insert size reduced total error during variant calling

Fig. 4

A Benchmarking F1 scores for SNPs and INDELs across all regions of the genome of the 3 insert size libraries. B Benchmarking F1 scores for SNPs and INDELs of difficult regions of the genome. C Benchmarking F1 scores for SNPs and INDELs of regions with low mappability and segmental duplications

Fig. 5

Median aligned insert size of PCR Free and PCR Plus libraries generated for Projects 1 and 2. The first 77 libraries for Project 1 were generated using the 350 bp insert method, with the remaining samples being prepared using the 600 bp insert or PCR Plus methods. The libraries with small aligned insert sizes indicate potential microbial contamination

Fig. 6

A Genomic coverage CV at 1 × coverage of 33 example samples. Samples with a CV greater than 0.1 were marked for additional screening and QC. B Example genomic coverage CV plot at 1 × of a sample with a CV of 0.09 yielding a “flat” coverage profile. C Non uniform genomic coverage profile of a sample with a CV of 0.14

Fig. 7

Total polonies per run over time of (A) 600 bp insert size Project 1 and (B) > 1 kb + insert library Project 2. Vertical dashed line 1 indicates adjustment of insert size to 600 bp. Vertical dashed line 2 indicates start of Bulk QC runs prior to trio sequencing

Fig. 8

Mean base quality by run over time of (A) 600 bp insert size Project 1 and (B) > 1 kb + insert library Project 2. Dashed horizontal lines indicate Q40 and Q30 quality scores

Fig. 9

A Mean base quality by run for each rWGS flow cell sequencing run at 2 × 100 and standard 2 × 100 Cloudbreak sequencing run. B Percent Q30 quality per Run. C Total benchmarking errors for HG001 when sequenced using either rWGS (40,267) and standard Cloudbreak (36,207) sequencing chemistry. Data was down sampled to 30 × using 2 × 100 read lengths for analysis

Fig. 10

Run metrics for the control Cloudbreak Freestyle (FS) 2 × 150 and 24 h rWGS 2 × 150. A Mean base quality, B Percent Q30 per run and C Benchmarking error of HG002 down sampled to 35 × coverage

Fig. 11

Kraken analysis of human samples potentially contaminated with bacterial species