ctDNA-based liquid biopsy technology holds great potential for facilitating the advancement of personalized cancer treatment. Though many ctDNA assay technologies have entered the market in recent years, validating the sensitivity and specificity of these technologies to assess their comparative performance has been challenging. This can largely be attributed to the lack of an accurate ctDNA reference at commonly assayed cancer hot spots. Current cfDNA standards have been derived from sonicated gDNA or synthetic oligonucleotides, neither of which is reflective of naturally-occurring ctDNA with regard to fragment size distribution and fragment end behavior.
Here we report the development of a cfDNA simulate containing common cancer hot spots which more closely resembles naturally-occurring cfDNA with respect to fragment size distribution, end behavior (ligation efficiency), and complexity. Our cfDNA simulate was generated to achieve fragments of 130-230bp and demonstrated a ligation efficiency comparable to that of human cfDNA (>70%). Finally, our cfDNA simulate set contains an array of single-nucleotide variations or deletions in common hot spots genes including: EGFR, KRAS, NRAS, PI3K, BRAF, among others. Allele frequencies at each hot spot were validated on a diluted panel (0%, 0.1%, 0.5%, 1%) using ddPCR and our Firefly-NGS assay. Our Firefly-NGS assay was able to detect all hot spot variants in a 40ng sample at 0.1% concentration.