Table 4.
|ΔΔG exp| < 1.37 | 1.37 ≤ |ΔΔG exp| < 2.73 | |ΔΔG exp| ≥ 2.73 | Total | |
---|---|---|---|---|
Wang et al.,12 OPLS2.1, Figure 3a | ||||
No. of transformations | 257 | 66 | 7 | 330 |
Underestimations [%] | 51 | 71 | 86 | 55 |
(ΔΔG cal–ΔΔG exp)/N | −0.04 | −0.59 | −1.56 | −0.18 |
Roos et al.,19 OPLS3, Figure 3c | ||||
No. of transformations | 348 | 98 | 23 | 469 |
Underestimations [%] | 55 | 71 | 74 | 59 |
(ΔΔG cal–ΔΔG exp)/N | −0.12 | −0.71 | −0.82 | −0.28 |
Roos et al.,19 OPLS3e, Figure 3d | ||||
No. of transformations | 348 | 98 | 23 | 469 |
Underestimations [%] | 52 | 71 | 65 | 57 |
(ΔΔG cal–ΔΔG exp)/N | −0.03 | −0.53 | −0.51 | −0.16 |
Pérez‐Benito et al.,20 all ligand pairs with 5 ns simulations from LO datasets, Figure 3b | ||||
No. of transformations | 10 | 79 | 92 | 181 |
Underestimations [%] | 70 | 65 | 83 | 74 |
(ΔΔG cal–ΔΔG exp)/N | −0.13 | −0.43 | −1.14 | −0.78 |
“Underestimations” is the percentage of pairs of which ΔΔG cal < ΔΔG exp when the ligand pairs are rearranged so that ΔΔG exp ≥ 0 kcal mol−1. The average underestimation (ΔΔG cal – ΔΔG exp)/N is also calculated after making the ΔΔG exp ≥ 0 rearrangement.