High-Quality Genome Sequencing With PCR-Free Library Preparation

PCR-free whole genome sequencing: A survey of enzymatic fragmentation DNA library preparation solutions Ariele Hanek, Philip Benson, Josh Haimes, Thomas Harrison, Andrew Gaines, Doug Wendel, Kristina Giorda, Martin Ranik, Brian Kudlow – Watchmaker Genomics, Boulder, CO For Research Use Only. Not for use in diagnostic procedures. © 2023 Watchmaker Genomics. All trademarks are the property of their respective owners. Introduction PCR-free library preparation is considered the gold standard for whole-genome sequencing with minimal bias. These libraries should contain appropriate insert lengths to make efficient use of 2 x 150 bp read structure; however, size selection must not be so stringent as to discard a high fraction of yield, as clinical samples are often scarce. This study describes the ideal parameters for PCR-free WGS library preparation using eight commercially available library prep solutions, and assesses the sequencing data quality of each one. Library Size Tuning Matching final library sizes was essential to make valid comparisons between the surveyed library preparation kits. Library size was first tuned by adjusting fragmentation temperature and time, with a target of 450 bp mode insert length. Post-ligation SPRI ratio was the second lever used to adjust library size, if minimal fragmentation parameters created libraries smaller than the desired length. Table 1. Input mass and post-ligation SPRI ratios 316 337 351 350 345 357 361 372 395 429 500 400 200 300 100 0 Mean Insert Size (bp) Watchmaker dsSPRI Watchmaker 75 ng KAPA HyperPrep KAPA HyperPlus KAPA EvoPlus NEBNext Ultra II FS IDT xGen EZ QIAseq FX Illumina Watchmaker Mean Insert Sizes Were Matched Between Kits 1.5 1.0 0.5 0 Count (Millions) Insert Size Distribution Insert Size (bp) 0 200 400 600 800 NEBNext Ultra II FS IDT xGen EZ QIAseq FX Illumina KAPA EvoPlus Watchmaker Watchmaker dsSPRI Watchmaker 75 ng KAPA HyperPrep KAPA HyperPlus 15 Size (bp) 5000 4000 3000 2000 1000 0 Sample Intensity (Normalized FU) 100 250 400 1000 600 1500 2500 3500 5000 10000 NEBNext Ultra II FS IDT xGen EZ QIAseq FX Illumina KAPA EvoPlus Watchmaker Watchmaker dsSPRI Watchmaker 75 ng KAPA HyperPrep KAPA HyperPlus Library Prep Kit Input Mass Post-ligation SPRI Watchmaker 300 ng 0.5x Watchmaker dsSPRI 300 ng 0.3 – 0.5x Watchmaker 75 ng 75 ng 0.5x KAPA HyperPrep 300 ng 0.3x KAPA HyperPlus 300 ng 0.1 – 0.25x KAPA EvoPlus 300 ng 0.4x NEB Ultra II FS 300 ng 0.25 – 0.35x IDT xGen EZ 300 ng 0.25x QIAseq FX 300 ng 0.25x Illumina 300 ng 0.8 – 1.8x A B C Figure 2. Fragmentation and SPRI cleanup parameters were tuned to match final library sizes. Libraries within 50 bp of the target 450 bp mode insert length were selected for sequencing. (A) TapeStation electropherogram traces of sequenced libraries. (B) Mean insert sizes were calculated and (C) insert size distributions were plotted using NovaSeq 6000 output data. Watchmaker Prep Kits Maximize PCR-free Yields The necessary input mass for PCR-free library preparation to produce sufficient yield for NovaSeq 6000 sequencing was determined. A minimum input of 300 ng resulted in approximately 4 nM final libraries across various chemistries. To demonstrate Watchmaker’s efficiency, libraries were also generated with just 75 ng of input, delivering superior yield and similar library insert size distribution to other methods (Figure 2C), as measured by qPCR. Library prep time: 1 hr 30 min / Hands-on time: 45 min 5X Frag/AT Master Mix Adapter 5 min @ 30ºC 30 min @ 65ºC 4ºC HOLD Frag/AT Buffer Frag/AT Enzyme Mix Ligation Master Mix Fragmentation and A-tailing Ligation 15 min @ 20ºC 0.5X SPRI Cleanup 300 ng input DNA WGS recommendations WGS recommendations Figure 1. PCR-free library prep recommended workflow. The Watchmaker DNA Library Prep Kit with Fragmentation offers a streamlined and automation-friendly workflow. Conclusions Watchmaker DNA Library Prep Kit with Fragmentation delivers: • High conversion of gDNA inputs without the need for PCR amplification • High quality, uniform and unbiased whole genome sequencing results • Best-in-class coverage across challenging GC-rich regions Sequencing Results Table 2. Sequencing coverage 2 1 0 Normalized Coverage Minimal GC-bias GC Content 0 25 50 75 100 NEBNext Ultra II FS IDT xGen EZ QIAseq FX Illumina KAPA EvoPlus Watchmaker Watchmaker dsSPRI Human Genome Watchmaker 75 ng KAPA HyperPrep KAPA HyperPlus Increased Coverage Across Challenging Promoters 35 25 15 30 20 10 Mean Coverage Watchmaker dsSPRI Watchmaker 75 ng KAPA HyperPrep KAPA HyperPlus KAPA EvoPlus NEBNext Ultra II FS IDT xGen EZ QIAseq FX Illumina Watchmaker High Quality Read Alignment 92.2 91.8 91.4 92.0 91.6 91.2 91.0 PF HQ Aligned Reads (%) Watchmaker dsSPRI Watchmaker 75 ng KAPA HyperPrep KAPA HyperPlus KAPA EvoPlus NEBNext Ultra II FS IDT xGen EZ QIAseq FX Illumina Watchmaker Library Prep Kit Mean Coverage Median Coverage Watchmaker 28.6 30 Watchmaker dsSPRI 28.8 31 Watchmaker 75 ng 28.2 30 KAPA HyperPrep 27.9 30 KAPA HyperPlus 28.9 31 KAPA EvoPlus 25.9 27 NEB Ultra II FS 28.1 30 IDT xGen EZ 29.0 31 QIAseq FX 28.6 30 Illumina 29.2 31 A B C Figure 4. Maximize coverage while minimizing biases. At whole genome sequencing depths (Table 2), Watchmaker libraries achieve (A) efficient alignment of high-quality reads, (B) unbiased-GC coverage and (C) exceptional coverage at GC-rich promoter regions.1 Low Chimeric Read Counts 0.030 0.020 0.025 0.015 0.010 0.005 0 Chimeras (%) Watchmaker dsSPRI Watchmaker 75 ng KAPA HyperPrep KAPA HyperPlus KAPA EvoPlus NEBNext Ultra II FS IDT xGen EZ QIAseq FX Illumina Watchmaker Robust Variant Calling 0.984 0.980 0.976 0.982 0.978 0.974 SNP Indel 0.973 0.970 F1-Score NEBNext Ultra II FS IDT xGen EZ QIAseq FX Illumina KAPA EvoPlus Watchmaker Watchmaker dsSPRI Watchmaker 75 ng KAPA HyperPrep KAPA HyperPlus Reduced Hairpin Artifacts 4000 3000 1500 2000 3500 2500 1000 500 0 Average Artifacts/Millio n Watchmaker dsSPRI Watchmaker 75 ng KAPA HyperPrep KAPA HyperPlus KAPA EvoPlus NEBNext Ultra II FS IDT xGen EZ QIAseq FX Illumina Watchmaker Final sequencing read P5 P5 A T 5' P7 P7 Softclip bases Single-stranded DNA annealing End repair + A-tailing Adapter ligation A B C D Figure 5. Prevent sequencing artifacts to increase the accuracy of variant calls. (A) A hypothesized mechanism of hairpin artifact formation, which interferes with variant calling accuracy. The rate of (B) hairpin artifacts and (C) chimeric reads in Watchmaker libraries were consistently some of the lowest observed. (D) The combination of low artifactual representation and high depth of coverage in Watchmaker libraries resulted in high precision and recall when calling SNP and indel variants. Methods PCR-free WGS. Enzymatic fragmentation libraries were constructed from 300 ng gDNA (NA12878, Coriell Institute) using the Watchmaker DNA Library Prep Kit with Fragmentation, KAPA HyperPlus, KAPA EvoPlus, NEBNext Ultra II FS DNA Library Preparation Kit, IDT xGen DNA Library Prep EZ Kit, QIAseq FX DNA Library Kit, or Illumina DNA PCR-free Prep Kit, per manufacturer’s recommendations. The sonication control library was constructed from 300 ng of Covaris sheared gDNA (NA12878, Coriell Institute) using the KAPA HyperPrep Kit. Watchmaker’s DNA Library Prep Kit with Fragmentation utilized a 5 minute at 30°C fragmentation followed by a 0.5X post-ligation SPRI for longer fragments. The post-ligation SPRI ratio for all other library prep kits was optimized to make 450 bp fragments as measured by HSD5000 ScreenTapes. Figure 3. Efficient conversion in high-stringency applications. Double-sided size selection was necessary to achieve matching final library insert sizes for some of the methods assessed. However, it’s important to note that this additional step could potentially have a detrimental effect on final library yields, while also extending the total workflow time and cost. Post-ligation library yields quantified by qPCR. High Yields Without Amplification 12 8 4 10 6 2 0 Yield (nM) Watchmaker dsSPRI Watchmaker 75 ng KAPA HyperPrep KAPA HyperPlus KAPA EvoPlus NEBNext Ultra II FS IDT xGen EZ QIAseq FX Illumina Watchmaker 1 Michael G Ross, Carsten Russ, Maura Costello, Andrew Hollinger, Niall J Lennon, Ryan Hegarty, Chad Nusbaum, David B Jaffe, Characterizing and measuring bias in sequence data, Genome Biology, 14(5):R51. doi: 10.1186/gb-2013-14-5-r51 M300