Biological systems rely on the establishment of interactions between biomolecules, which take place in the aqueous environment of the cell. It was already demonstrated that a small set of residues at the interface, Hot-Spots(HS), contributes significantly to the binding free energy. However, these energetic determinants of affinity and specificity are still not fully understood. Moreover, the contribution of water to their HS character is also poorly characterized. In this review, we have focused on the structural data available that support the occlusion of HS from solvent, and therefore the "O-ring theory"not only on protein-protein but also on protein-DNA complexes. We also emphasized the use of Solvent Accessible Surface Area (SASA) features in a variety of machine-learning approaches that aim to detect binding HS.