Fragment Based Molecular Dynamics for Drug Design Lucia Sessa 1(&) , Luigi Di Biasi 1 , Simona Concilio 2 , and Stefano Piotto 1 1 Department of Pharmacy, University of Salerno, Via Giovanni Paolo II, 132, 84084 Fisciano, SA, Italy lucsessa@unisa.it 2 Department of Industrial Engineering, University of Salerno, Via Giovanni Paolo II, 132, 84084 Fisciano, SA, Italy Abstract. Molecular docking is a computationally efficient method used to predict the conformations adopted by the ligand within a target-binding site. A positive aspect of conventional docking is the possibility of easily distributing the calculation on dedicated grid or cluster. The receptor is usually kept rigid, therefore the changes in the binding pocket geometry induced by the ligand is overlooked. Here we present a new docking approach (DynDock) that exploits molecular dynamics to preserve the flexibility of the receptor. To maintain high computational efficiency, DynDock has been developed to be distributed on a grid. The main advantages of this method are the full flexible molecular docking achieved during the simulation and the reduced number of compounds collected. Keywords: Docking  Drug design  Molecular dynamics 1 Introduction The molecular design is a computationally demanding task; it is the process of finding new drugs and involves the design of molecules that are complementary to the target in shape and charge. Usually, these compounds interact with a protein activating or inhibiting its function. There are two major methods of molecular design. The first is the Ligand-Based Drug Design (LBDD) that uses the structural characteristics of all molecules that bind the target of interest, to derive a pharmacophore model [1]. The second method is the Structure-Based Drug Design (SBDD), which is based on knowledge of the three-dimensional structure of the target [2]. The aim is to predict the affinity and the selectivity of a drug candidate using the ligand and the target structure. In details, SBDD is a cyclic process, which starts from a known target structure usually experimentally obtained by X-ray crystallography or NMR spectroscopy [3]. The knowledge of 3D structures permits to run in silico studies to identify potential ligands (Fig. 1). Following the molecular modelling predictions, the most promising compounds can be synthesized and evaluated for their biological properties. Once synthesized and © Springer International Publishing AG, part of Springer Nature 2018 M. Pelillo et al. (Eds.): WIVACE 2017, CCIS 830, pp. 49–58, 2018. https://doi.org/10.1007/978-3-319-78658-2_4