Fig. 6: Distributions of in silico binding free energies of PXN (1) and 30 to α1β3γ2L GABA_A and RDL.

Computed binding free energies predict that 30 is not as potent as PXN (1) toward the GABA_A receptor (ΔΔG = +1.8 ± 1.5), whereas 30 is much more potent than PXN (1) toward the RDL receptor (ΔΔG = −3.9 ± 1.6), in agreement with experimental data. A standard, two-tailed T-test showed statistical significance in the differences between the binding free energies of PXN (1) and 30 for the GABA_A receptor (p = 2.95×10⁻³⁰) and the RDL receptor (p = 1.22×10⁻³¹). Each distribution except for 30@GABA_A passed at least one test of normality (Shapiro–Wilk, D’Agostino’s K2, Anderson-Darling). Small dotted lines indicate the interquartile range of predicted binding free energies, while the thick dotted line indicates the 50th percentile.