Establishing community reference samples, data and call sets for benchmarking cancer mutation detection using whole-genome sequencing

Whole genome and exome sequencing reference datasets from a multi-center and cross-platform benchmark study

“…In addition, we inferred subclones and heterogeneity of HCC1395 with bulk DNA sequencing. The results were confirmed by single-cell DNA sequencing analysis 3 .…”

“…The rapid growing number of sample processing protocols, library preparation methods, sequencing platforms, and bioinformatics pipelines to detect mutations in cancer genome, presents great technical challenges for the accuracy and reproducibility of utilizing NGS for cancer genome mutation detections. To investigate how these experimental and analytical elements may affect mutation detection accuracy, recently we carried out a comprehensive benchmarking study 2 using both whole-genome (WGS) and whole-exome sequencing (WES) data sets generated from two well-characterized reference samples: a human breast cancer cell line (HCC1395) and a B lymphocytes cell line (HCC1395BL) derived from the same donor 3 . We generated WGS and WES data using various NGS library preparation protocols, seven NGS platforms (NovaSeq, HiSeq, PacBio, 10X Genomics, Ion Torrent, Miseq, and Affymetrix CytoScan HD) at six centers including Illumina (IL), National Cancer Institute (NC), Novartis (NV), European Infrastructure for Translational Medicine (EA), Fudan University (FD), and Loma Linda University (LL) (Fig.…”

“…We first established reference call sets with evidence from 21 replicates of Illumina WGS runs with coverage ranging from 50X to 100X (1150X in total). We split mutation call confidence levels into four categories: HighConf, MedConf, LowConf, and Unclassified 3 . By combining all WGS runs, we were able to further confirm and improve our call set with tumor-normal pairs of 1500X data sets and identified mutations with VAF as low as 1.5%.…”

The Sequencing Quality Control 2 study: establishing community standards for sequencing in precision medicine

“…These materials enable scientists to establish and benchmark their NGS workflow, compare performance with consortium data [ 10 , 15 ], and guide efforts by related scientific communities, such as the Association of Biomolecular Resource Facilities (ABRF). Similarly, the large number of datasets and protocols generated during SEQC2 are available as an accessible resource for ongoing development of bioinformatic tools [ 12 ]. However, despite these resources, final clinical validation of NGS assays using patient samples is required prior to clinical use.…”

“…The cancer genome often harbors large and complex mutations that result from genome instability. WGS can be required to diagnose these complex mutations, including translocations, loss-of-heterozygosity, and gene amplifications and deletions [ 12 ]. The SEQC2 consortium performed WGS using the reference tumor samples to understand the variables that impact diagnoses.…”