Research Papers:

Gene-based comparative analysis of tools for estimating copy number alterations using whole-exome sequencing data

Hyung-Yong Kim _, Jin-Woo Choi, Jeong-Yeon Lee and Gu Kong

PDF  |  HTML  |  Supplementary Files  |  How to cite

Oncotarget. 2017; 8:27277-27285. https://doi.org/10.18632/oncotarget.15932

Metrics: PDF 2123 views  |   HTML 3196 views  |   ?  


Hyung-Yong Kim1, Jin-Woo Choi1, Jeong-Yeon Lee2, Gu Kong1,2

1Department of Pathology, College of Medicine, Hanyang University, Seoul, Republic of Korea

2Institute for Bioengineering and Biopharmaceutical Research (IBBR), Hanyang University, Seoul, Republic of Korea

Correspondence to:

Gu Kong, email: [email protected]

Keywords: cancer CNV, CNA estimation, WES, NGS, copy number

Received: April 10, 2016     Accepted: February 20, 2017     Published: March 06, 2017


Accurate detection of copy number alterations (CNAs) using next-generation sequencing technology is essential for the development and application of more precise medical treatments for human cancer. Here, we evaluated seven CNA estimation tools (ExomeCNV, CoNIFER, VarScan2, CODEX, ngCGH, saasCNV, and falcon) using whole-exome sequencing data from 419 breast cancer tumor-normal sample pairs from The Cancer Genome Atlas. Estimations generated using each tool were converted into gene-based copy numbers; concordance for gains and losses and the sensitivity and specificity of each tool were compared to validated copy numbers from a single nucleotide polymorphism reference array. The concordance and sensitivity of the tumor-normal pair methods for estimating CNAs (saasCNV, ExomeCNV, and VarScan2) were better than those of the tumor batch methods (CoNIFER and CODEX). SaasCNV had the highest gain and loss concordances (65.0%), sensitivity (69.4%), and specificity (89.1%) for estimating copy number gains or losses. These findings indicate that improved CNA detection algorithms are needed to more accurately interpret whole-exome sequencing results in human cancer.

Creative Commons License All site content, except where otherwise noted, is licensed under a Creative Commons Attribution 4.0 License.
PII: 15932