Tumour-guided Personalised Sequencing Assays For ctDNA Detection

Tumour-guided Personalised Sequencing Assays For ctDNA Detection

Abstract
Detecting extremely rare variant alleles is becoming increasingly relevant for developing novel diagnostic tools in diverse fields, such as oncology, prenatal diagnostics and transplantation medicine. Significant challenges in analysing cell-free DNA from blood plasma are limited amounts of material, highly degraded source material, and the presence of variants of interest at extremely low variant allele frequency. Here we present a highly optimised workflow for creating tumour-guided personalised ultrasensitive sequencing assays for cancer detection from liquid biopsies. This includes advanced in silico assay design software designed specifically for developing multiplex assays based on the SiMSen-seq technology, carefully optimised reaction chemistry and a machine-learning enhanced variant calling algorithm. These enhancements significantly increase the dynamic range and sensitivity to detect rare mutations and recover target molecules. We also provide tools for the analysis of sequencing data containing UMI. Thus, SiMSen-seq facilitates highly accurate variant detection in challenging sample types such as liquid biopsies. Lastly, we demonstrate assay performance using real-world data from clinically relevant material.