A somatic reference standard for cancer genome sequencing

Abstract

Large-scale multiplexed identification of somatic alterations in cancer has become feasible with next generation sequencing (NGS). However, calibration of NGS somatic analysis tools has been hampered by a lack of tumor/normal reference standards. We thus performed paired PCR-free whole genome sequencing of a matched metastatic melanoma cell line (COLO829) and normal across three lineages and across separate institutions, with independent library preparations, sequencing, and analysis. We generated mean mapped coverages of 99X for COLO829 and 103X for the paired normal across three institutions. Results were combined with previously generated data allowing for comparison to a fourth lineage on earlier NGS technology. Aggregate variant detection led to the identification of consensus variants, including key events that represent hallmark mutation types including amplified BRAF V600E, a CDK2NA small deletion, a 12 kb PTEN deletion, and a dinucleotide TERT promoter substitution. Overall, common events include >35,000 point mutations, 446 small insertion/deletions, and >6,000 genes affected by copy number changes. We present this reference to the community as an initial standard for enabling quantitative evaluation of somatic mutation pipelines across institutions.

Resolving sub-clonal heterogeneity within cell-line growths by single cell sequencing genomic DNA

“…3) exhibit extensive copy number changes. At a 2Mb resolution cell #596 has three distinct copy number states in chromosome 1 (3-2-4), two in chromosome 18 (2-3) and a single ploidy of 3 across chromosome 8; in contrast #415 exhibits two copy number states (2)(3)(4) in chromosome 1, while chromosome 18 is a single segment at copy number 2 and chromosome 8 is at a different copy number of 4 ( Supplementary Fig. 3).…”

“…The standard paradigm is bulk sequencing of genomic DNA derived from millions of heterogeneous cells. In bulk sequencing, the ability to resolve sub-clonality is largely lost except for indirect inference, resulting in an ensemble view dominated by the majority clone 1,2 . While bulk sequencing has provided major insights into tumor biology, low-resolution single cell methods, such as spectral karyotyping, show that dissecting events at single cell resolution is critical to accurately describe the genome of heterogeneous population of cells that underlie sub-clonal complexity and tumor evolution.…”

“…The melanoma COLO829 and germline COLO829-BL tumor/normal pair have been extensively analyzed using multiple methods and technologies [1][2][3][4] . The first exhaustive sequencing of this tumor/normal pair by Pleasance et al described several hallmark events including a homozygous 12kb deletion in PTEN, BRAF 600V/E, and a CDK2NA 2bp deletion.…”

“…The first exhaustive sequencing of this tumor/normal pair by Pleasance et al described several hallmark events including a homozygous 12kb deletion in PTEN, BRAF 600V/E, and a CDK2NA 2bp deletion. Previous studies using bulk sequencing in the tumor-line COLO829 have focused largely on developing tools, catalogues and standardizations to improve copy number estimation and cancer characterization 2 . While a few of the studies found cell line complexity inconsistent with the assumption of clonality and suggestive of multiple sub-clones, in general, most analyses presumed COLO829 to be a single clone.…”

Establishing reference samples for detection of somatic mutations and germline variants with NGS technologies

“…Previous attempt has characterized a cancer cell line (from metastatic melanoma) that 111 inquired somatic mutations (SNV/indels) in exon regions only. Germline variants and somatic 112 mutations across the rest of the genome were not defined 11 . In addition, this dataset is 113 distributed under dbGAP-controlled access, limiting its accesibility and utility.…”