Fig. 6
From: Nonlinear mixed-effects modelling for single cell estimation: when, why, and how to use it
Noise estimation for Model 1. Analysis in the case of unknown variance of the measurement noise. The noise is assumed to be additive and the same in all cells. For STS, one estimate of σ is obtained from each cell (blue bars). a For Rich Data, STS can estimate noise through the average of the values (red line) equally good as NONMEM (black line), even though some individual cells display worse estimates. b For the case of Sparse Data, however, the mean from STS is significantly worse than for NONMEM