Fig. 2: Validation of the ecDNA amplicons using Circle-seq.

a Summary of ecDNA overlapping lists from the prediction of AmpliconArchitect (AA) and identification using Circle-seq. The y-axis is the ecDNA amplicon number from WGS prediction. Overlap: the ecDNA amplicons were identified using both AA software from WGS and Circle-Seq. None: the ecDNA amplicons were only identified using AA software but not using Circle-Seq. b The genome browser track at the FGFR2 gene locus from whole-genome sequencing (WGS) and Circle-seq. The connection lines on the top represent the potential structure combination in ecDNA amplicons predicted by AA software. N normal tissue, T tumor tissue. c Circular DNA elements identified from Cricle-Seq were separated into ecDNA (copy number ≥ 7) (The number of ecDNA, n = 0, n = 41, n = 42, n = 216, n = 0, n = 0, n = 714, n = 261, n = 0, n = 222 for S21, S25, S26, S27, S28, S29, S30, S31, S32 and S33, respectively) and extrachromosomal circular DNA (eccDNA) (copy number < 7) (The number of eccDNA, n = 1757, n = 7264, n = 4813, n = 38,804, n = 3610, n = 572, n = 2177, n = 3420, n = 481, n = 1686 for S21, S25, S26, S27, S28, S29, S30, S31, S32 and S33, respectively). The box plots show the minima (bottom dot), the maxima (top dot), the median (middle line) and the first and third quartiles (boxes), whereas the whiskers show 1.5× the interquartile range IQR above and below the box. Source data are provided as a Source Data file for Fig. 2a–c.