This article is focused on the application of two types of docking software, AutoDock and DOCK. It is aimed at studying the interactions of a calcium-dependent bacterial lectin PA-IIL (from Pseudomonas aeruginosa) and its in silico mutants with saccharide ligands. The effect of different partial charges assigned to the calcium ions was tested and evaluated in terms of the best agreement with the crystal structure. The results of DOCK were further optimized by molecular dynamics and rescored using AMBER. For both software, the agreement of the docked structures and the provided binding energies were evaluated in terms of prediction accuracy. This was carried out by comparing the computed results to the crystal structures and experimentally determined binding energies, respectively. The performance of both docking software applied on a studied problem was evaluated as well. The molecular docking methods proved efficient in identifying the correct binding modes in terms of geometry and partially also in predicting the preference changes caused by mutation. Obtaining a reasonable in silico method for the prediction of lectin-saccharide interactions may be possible in the future.