Post-Docking Optimization and Analysis of Protein-Ligand Interactions of Estrogen Receptor Alpha using AMMOS Software

摘要

In silico approaches have gained immense significance in the recent years and have become an integral part of research in both, academic institutions and different industry branches, thus assisting drug design and discovery. Among these methods the virtual ligand screening (VLS) of bioactive compounds has been established as an effective approach to handle large sets of compounds improving in this way the "hit-rate" of drug discovery programs [1-3]. Structure-based VLS (SB-VLS) relies on the three dimensional structure of the biological target obtained either through experimental methods such as X-ray crystallography or NMR spectroscopy [4], or predicted by homology modeling. Most of the SB-VLS protocols [5-7] employ docking of all compounds in an appropriate chemical library into the binding pocket of the selected target and evaluate the fit between the molecules via scoring functions. Critical for docking accuracy is the treatment of the ligand flexibility and, in many cases, the receptor flexibility [8-10], the last increasing considerably the computational time. It is impractical to perform SB-VLS using docking on a fully flexible protein receptor for a large number of ligands. It has been shown that post-docking optimization, either after conventional docking-scoring procedures or after hierarchical VLS protocols [6, 11, 12] helps in improving both, the docking pose and the scoring, and thus the overall efficiency of SB-VLS. This view is supported by a number of examples of binding pose prediction and enrichment improvements after post-docking energy minimization [13-17].