Complex noise-vs-mean dependency of single-cell phenotypes. Each plot represents 185 samples (5 per strain) of at least 200 cells each. x-axis: ‘mean’ value of the trait of interest in the sample. y-axis: coefficient of variation of the trait values within the sample. Four traits (indicated in headers) are shown as examples to indicate that the noise vs. mean dependency can be linear (A and C) or not (B and D), and negative (A) or positive (C). Red line: Lowess regression model used to condition noise on mean values. D155_A: Angle between [from cell center to brightest point in nucleus] and [end of major axis of the cell] at G1. A7-1_A: Size of actin region at G1. D182_A: Nuclear axis ratio at G1. C104_A: Short axis length in whole cell at G1.