HIV-1プロテアーゼ, P38, CDK-2複合体の立体構造を用いたドッキングソフトウェアの能力評価

概要

論文の詳細を見る
In this study, we evaluated the docking accuracy of software ArgusLab, which is one of freely available computational docking programs both for commercial and academic users, by using experimentally determined protein-ligand complex structures. The complexes include one of three proteins, HIV-1 protease, p38 and CDK-2, as target proteins. Because these proteins are well known drug targets, the tests can be suitable for practical use. All three steps of computational docking, i.e. pose construction, pose selection, and virtual screening, were tested for 16 complexes. The results indicate that ArgusLab can generate reasonable complex structures for around 50% of protein-ligand systems, and ligand flexibility play important roles in docking accuracy. Although pose construction and pose selection were successfully carried out both for HIV-1 protease and for CDK-2 complexes, for binding free energy calculations, which is one of the important steps in virtual screening, the accuracy of results for HIV-1 protease complexes was better than those of kinase complexes. By using these results as reference, reliable docking calculations are expected to be carried out by ArgusLab.
東北薬科大学の論文