Fig. 4

Relationship between insecticide choice and geographic location on probability of resistance and time to first-to-break with nuclear-only inheritance (NN). Insecticides differ by their effectiveness and geographies differ by their female exposure. The probability of resistance describes the fraction of all simulated runs where the data type is ‘Successful Measurement’ or ‘Toward Threshold’ (see Fig. 1). The time to first-to-break is calculated from the ‘Successful Measurement’ data type only. In each panel, the bold-colour lines (per strategy purple-to-yellow; R:viridis) come from partitioning the y-axis parameter from the simulations by the x-axis parameter into 101 rounded bins and calculating the mean of the y-axis measure per bin; a \(k=5\) backward-tail moving average is used to smooth the mean-line. Around each bold-colour line, there is a transparent-shading of the same colour that describes the 95% confidence intervals for the mean (\(\pm 1.96*SE\)), which is also smoothed with a \(k=5\) backward-tail moving average. In panels A and D, the x-axis is arbitrarily designated for a focal insecticide as effectiveness 1, as if it were a new insecticide. In panels B and E, effectiveness 1 is assumed to be > 0.8 (in accordance with WHO guidelines for new ITNs), and the x-axis is then for a partner insecticide. In panels C and F, effectiveness 1 is also assumed to be > 0.8, and the x-axis is then for fem ale exposure