Abstract 1307: Effect of tumor purity on somatic mutation detection using next-generation sequencing technology: A benchmarking study

Abstract

Abstract Next generation sequencing (NGS) is becoming an essential tool for clinical cancer care. Whole genome sequencing (WGS), whole exome sequencing (WES), and targeted gene sequencing (TGS) have emerged as important tools and are critical for advancing precision oncology to meaningfully impact patient lives. To enable this goal, we must first thoroughly characterize the complexities of using NGS to identify somatic mutations in samples with tumor heterogeneity and variable tissue sample purity. Here, we applied both experiment and in silico approaches to determine the effects of tumor heterogeneity on somatic mutation detection. First, we systematically investigated somatic mutation events in paired cell lines (breast carcinoma and matched normal) using a myriad of NGS platforms and protocols. Then, we applied machine learning algorithms to define high confidence mutation calls with datasets from the combination of all platforms and pipelines. Finally, we performed whole genome sequencing from tumor DNA samples titrated with matched normal cells to create a gradient of tumor purities from 100%, 75%, 50%, 20%, 10%, 5%, and 0%. We also created a series of in silico mixes by pooling reads from tumor and normal cell lines at respective ratios. With somatic variation “ground truth” established by multiple orthogonal sequencing platforms and multiple analysis tools, we evaluated performance of individual software tools for tumor purity assessment and for mutation detection. Our experimental design allowed us to systematically evaluate accuracy, specificity and sensitivity of mutation detection across a range of tumor purities, as well as conduct a head-to-head comparison of in silico approaches versus experimental approaches. This comprehensive study will help to clarify our understanding of experimental and analytical challenges that arise when confronted with tumor heterogeneity in patient care clinical decisions. In addition, data generated in this study will be used as a benchmark resource for the community to develop and improve tools to meet clinical challenges. Citation Format: Wendell Jones, Wenming Xiao, Somatic Mutation Working Group, SEQC2 Consortium. Effect of tumor purity on somatic mutation detection using next-generation sequencing technology: A benchmarking study [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2018; 2018 Apr 14-18; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2018;78(13 Suppl):Abstract nr 1307.