Fig. 4: Lift curves of the winning model versus the random (red) and base models (cyan).
From: Statistical modeling of SARS-CoV-2 substitution processes: predicting the next variant
We compare the winning model, the third-ranked Poisson model of non-synonymous amino acid substitutions, against two baselines. The first is the random ordering of all possible relevant substitutions (red), and the second is a base model, which considers the exposure and the transition/transversion (ti/tv) ratio, but not the other explanatory factors (cyan). To compare, we show the lift score (the ratio of true positives compared to a baseline model) as a function of the proportion of states considered. Source data: supplementary data 3.
