Cell-free DNA (cfDNA) analysis can allow clinicians to gain information regarding the genetic state and overall functioning of tissues within the body, both in normal and tumorigenic states. When apoptosis occurs in varying blood cell lineages, cfDNA can be detected within the blood. In patients with cancer, a small fraction of the cfDNA known as circulating tumor DNA (ctDNA) originates from dying tumor cells. As a non-invasive means of monitoring both disease progression and treatment efficacy, cfDNA variant calling has emerged recently as the next-generation sequencing (NGS) method.
Claret Bioscience has developed an easy and efficient next-generation sequencing (NGS) preparation method called single reaction single-stranded DNA library (SRSLY) that features an optional unique molecular identifier (UMI) add-on step for accurate detection of rare ctDNA variants. UMIs are used to uniquely tag each molecule in a sequencing library, acting as short molecular barcodes to aid in the detection of single nucleotide polymorphisms (SNPs) and resolve polymerase chain reaction (PCR) duplicates. In the analysis of NGS data from cancer samples, true pathogenic variant alleles are distinguished from variants introduced by copying errors anywhere in the library preparation and sequencing process. SRSLY is a unique NGS library prep kit that renders all input DNA single-stranded before adapter ligation and captures all DNA strands even when nicked or damaged.
Advantages of SRSLY
SRSLY library preparation can be used to detect common cancer variants at low allele fractions from varying sample input amounts. The detection of low allele frequency variants with known variant allele frequency (VAF) is performed using Horizon Discovery’s OncoSpan cfDNA standard, a well-characterized genomic reagent, and Twist Bioscience’s custom Pan Cancer targeted-enrichment Panel. Compared to traditional double-stranded DNA library preparations, SRSLY exhibits increased library complexity and similar UMI error rate and VAF calling sensitivity at all input amounts tested.
How OncoSpan reference standard adds value
The Horizon OncoSpan cfDNA standard is a well-characterized cell line-derived reference standard containing over 375 variants from over 152 key cancer genes present at between 1 and 100% variant allele fraction. An assessment using Horizon OncoSpan cfDNA with Claret Bioscience’s SRSLY Library Prep kit with UMI addition was performed to evaluate performance in ctDNA variant analysis. The libraries were sequenced and processed by the bioinformatics pipeline with the estimated library size equivalent to the estimated number of unique molecules present in the library. SRSLY showed a marked increase at higher inputs, with approximately double the estimated library size of the DS prep at multiple levels of input.
Reads per UMI are defined as the number of reads sequenced containing a given UMI, while the count is the number of UMIs with that number of reads per UMI, and a lower number of reads for each given UMI is preferred. At lower inputs, SRSLY and DS preps have a higher proportion of UMIs with multiple reads per UMI, illustrating that higher cfDNA inputs are ideal for variant calling. For all inputs tested, SRSLY has a higher number of UMIs with a single read, and at a lower maximum number of reads due to increased library complexity. Mean error rate was determined for each UMI that had three or more reads. The mean error rate was determined by first calling consensus sequences from reads with the same UMI, and then determining the error rate of each read from that consensus sequence. At lower inputs, the mean error rates are roughly similar between the two, however at the higher inputs SRSLY has a lower error rate than DS.
SRSLY’s UMI add-on protocol demonstrates excellent performance in variant detection applications. Compared to traditional library preparations, SRSLY libraries have slightly higher complexity and better polymerase fidelity during PCR at higher inputs and provide similar variant calling performance at identical sequencing depths. These data also demonstrate that the OncoSpan cfDNA can be used as standard to evaluate and deploy SRSLY for cfDNA-based variant calling in research use and clinical settings.