The prediction of orthosteric, allosteric, and cryptic sites through computational approaches is significant for drug discovery studies. To this extent, a ligand binding site prediction method called Essential Site Scanning Analysis (ESSA) was used to investigate allosteric and cryptic sites of TEM-1 beta-lactamase, cell division protein kinase 2, and galactokinase by applying various cutoff values (7 Å, 10 Å, 11 Å, and 13 Å) and combining 10/20 modes in this study. On the other hand, molecular dynamics (MD) and ClustENMD were performed to investigate hGALK1 enzyme dynamics. In addition, Principal Component Analysis (PCA) was applied to investigate collective motions in the presence and the absence of an allosteric inhibitor in hGALK1. The ligand binding sites of hGALK1 were investigated using frames obtained from MD and ClustENMD. According to the findings, the application of combined 10/20 modes improved the success of the ESSA method. Moreover, binding site prediction success demonstrated variation in the presence of different cutoff values and conformations. Different cutoff values demonstrated success in the prediction of allosteric and orthosteric sites highlighting the role of contacting atoms in the Gaussian Network Model (GNM) calculations. To this extent, the combination of results from various cutoff values has the potential to improve the allosteric site prediction success of the ESSA method