highly significantp = 1.725 × 10 –12
To statistically evaluate the differences in binding energies, a Kruskal–Wallis test was conducted, yielding highly significant results for both GNINA ( p = 1.725 × 10 –12 ) and Autodock rescoring ( p = 1.514 × 10 –12 ), confirming that binding energy distributions across receptors are not equivalent.