Optimizing Accuracy and Efficiency in Analyzing Non-UMI Liquid Biopsy Datasets Using the Sentieon ctDNA Pipeline

Abstract

AbstractSequencing clinical liquid biopsy, especially circulating tumor DNA (ctDNA), provides a valuable method for identifying low allele frequency tumor variants, opening novel clinical applications, particularly in treatment selection for late-stage cancer patients. Despite advancements, challenges in assay development persist, primarily due to limited sample volumes and insufficiency of reads supporting low allele frequency variants. The allele frequencies of clinically significant variants often hover close to the threshold of errors introduced by PCR and sequencing processes. Therefore, more sophisticated analysis methods are crucial to further reduce base error rates, enabling accurate discrimination between background errors and genuine somatic variants. While several ctDNA analysis pipelines have been published and adopted, there is room for improvement in terms of accuracy and run efficiency.In this study, we introduce Sentieon’s innovative consensus-based ctDNA pipeline - a rapid and precise solution for calling small somatic variants from non-UMI ctDNA sequencing data. The pipeline comprises four core modules: alignment, consensus generation, variant calling, and variant filtering. Through benchmarking with in-vitro and real clinical datasets, we observed that the Sentieon ctDNA pipeline exhibits higher accuracy compared to alternative methods.