Figure 3.

Linear Regression of ${\mathit{F}}_{\mathit{1}}$-scores and recall based on observations versus predictions. Linear regression (red lines) of $F_1$-scores and recall – computed between observations and predictions – for each peak caller proves their linear relationship (adjusted R squared of 0.92, 0.84 and 0.79, respectively). Each dot represents a tuple of FDR (or z-score in case of F-Seq) cutoff, cell line and replicate.