Fig. 4

Results on breast cancer patient S0. a Allele-specific copy number profiles for CNRein, SIGNALS, CHISEL, and Alleloscope on \(n=785\) cells. b Allele-specific copy number profiles of chromosome 6, showing that while CNRein retains the ability to detect small CNAs also detected by SIGNALS. c All methods have a similar fit to the read depth data, with SIGNALS and CNRein having a slightly better fit. d The size of the 50 largest clones for each method, showing much larger clones for CNRein than SIGNALS, CHISEL SIGNALS, and Alleloscope. e CNRein’s predictions resulted in a tree with a far lower parsimony score (397) than SIGNALS (40,222), CHISEL (8506), or Alleloscope (2076). f VAFs for copy number {2, 1} for CNRein, showing the expected peaks around 1/3 and 2/3. g Log likelihood ratios (LLR) of SNVs support on bootstrapped replicates between our method and the alternative methods SIGNALS, CHISEL, and Alleloscope. CNRein outperforms CHISEL (LLR 1093.83 and \(p < 10^{-5}\) ) and somewhat outperforms SIGNALS (LLR 145.26 and \(p=0.0062\)) while being slightly outperformed by Alleloscope (LLR \(-35.24\) and \(p=0.205\))