Figure 2: Experimental assessment of deepSNV.
From: Reliable detection of subclonal single-nucleotide variants in tumour cell populations
(a) Observed frequency distributions for SNVs and sequencing errors in a test data set of known composition. The dashed line shows the expected frequency based on an additive error model. (b) Scatter plot of nucleotide frequencies in control and mixture sample, and indicated levels of significance of the deepSNV test. The null hypothesis of the test implies that the observed relative frequencies lie close to the diagonal, whereas the alternative allows the frequency in the test experiment to be greater than in the control. (c) Receiver-operating characteristic curves demonstrate high power at low false-positive rates.
