1. Technical Validation of a Next-Generation Sequencing Assay for Detecting Clinically Relevant Levels of Breast Cancer–Related Single-Nucleotide Variants and Copy Number Variants Using Simulated Cell-Free DNA 
Xin Yang, Yuxing Chu, Rui Zhang, Yanxi Han, Lucheng Zhang, Yu Fu, Dan Li, Rongxue Peng, Dongdong Li, Jiansheng Ding, Ziyang Li, Meiru Zhao, Kuo Zhang, Tian Lu, Lang Yi, Qisheng Wu, Guigao Lin, Jiehong Xie, Tao Liu, Ling Yang, Xin Yi, Jinming Li 
The Journal of Molecular Diagnostics 
Volume 19, Issue 4, Pages (July 2017)
DOI: /j.jmoldx
Copyright © 2017 American Society for Investigative Pathology and the Association for Molecular Pathology Terms and Conditions

2. Figure 1
Comparison of plasma cell-free DNA (cfDNA) and mimic cfDNA quality metrics. A: Average on-target sequencing depth without duplicated reads (×). B: G + C content (%). C: Fraction of on-target effective bases without duplicated reads (capture efficiency, %). D: Fraction of uniquely mapped on-target reads (%). E: Fraction of reads mapped to genome (mapping rate, %). F: Fraction of target coverage >30× (%). G: Fraction of target coverage ≥500× (%).
The Journal of Molecular Diagnostics  , DOI: ( /j.jmoldx )
Copyright © 2017 American Society for Investigative Pathology and the Association for Molecular Pathology Terms and Conditions

3. Figure 2
Detection performances. The sensitivity, specificity, and accuracy of each mutation frequency were evaluated, respectively. With the increase of mutant allele frequency, the three indicators all increase, especially for the sensitivity. A: Single-nucleotide variant (SNV). B: Copy number variation (CNV).
The Journal of Molecular Diagnostics  , DOI: ( /j.jmoldx )
Copyright © 2017 American Society for Investigative Pathology and the Association for Molecular Pathology Terms and Conditions

4. Figure 3
Coverage as a function of single-nucleotide variant (SNV) sensitivity. Sensitivity was assessed as a function of coverage when reads were down-sampled to simulate average coverage of 200×, 600×, 1000×, 1400×, and 1800× in different mutant allele frequencies (shown by different symbols). As expected, sensitivity decreases with reduced coverage depth.
The Journal of Molecular Diagnostics  , DOI: ( /j.jmoldx )
Copyright © 2017 American Society for Investigative Pathology and the Association for Molecular Pathology Terms and Conditions

5. Supplemental Figure S1
Comparison of sequencing quality score. Quality scores across all bases: mimic cfDNA (A and C) and plasma cfDNA (B and D).
The Journal of Molecular Diagnostics  , DOI: ( /j.jmoldx )
Copyright © 2017 American Society for Investigative Pathology and the Association for Molecular Pathology Terms and Conditions

6. Supplemental Figure S2
Consistency of results between runs. A: Single-nucleotide variant. B: Copy number variation.
The Journal of Molecular Diagnostics  , DOI: ( /j.jmoldx )
Copyright © 2017 American Society for Investigative Pathology and the Association for Molecular Pathology Terms and Conditions

7. Supplemental Figure S3
Correlation analysis for digital PCR (dPCR) results with expected frequency and next-generation sequencing (NGS). A: The linear regression of correlation for dPCR results and expected frequency: R2 = 0.9016, slope = 0.9178, P <  B: The Pearson correlation coefficient for NGS and dPCR results: r = 0.9678, P < 
The Journal of Molecular Diagnostics  , DOI: ( /j.jmoldx )
Copyright © 2017 American Society for Investigative Pathology and the Association for Molecular Pathology Terms and Conditions