Fig. 6
From: Dose–response prediction for in-vitro drug combination datasets: a probabilistic approach
The figure shows observed versus predicted viability measurements for the baseline (Bliss) model (left panel), the PIICM (middle panel), and the comboLTR model (right panel). The points have been coloured according to the density of points. Overall, the performance of the three models are similar, with the main differences being how the models handle observations outside the [0, 1] interval. Both the baseline and PIICM are limited to prediction within this interval, while the comboLTR is not