Modeling the Relationship between Biological Activity of Delta-selective Enkephalin Analogues and Docking Results by Polynomials

Abstract

One of the areas of bioinformatics is to develop a fast and reliable method for predicting the biological activity of compounds. This will abbreviate the way for design of new compounds and reduce costs. The process of creating the selective ligands of delta opioid receptor (DOR) was directed towards the synthesis of enkephalin analogues. Their biological activity was determined using the in vivo and in vitro methods, allowing establishing the relationship between structure and biological activity. The application of computational methods in the design of this type of compounds reduces the stages of synthesis and biological tests. The relationship of the efficacy with the values of the so-called ChemScore scoring function from GOLD 5.2 and the values of total energy of ligand-receptor complex was modeled with first- to third-degree polynomials and surface fitted method. The polynomial surface of the third degree has the best fit, assessed by least squares method. In our previous study with theoretical model of DOR (PDBid:1ozc) was established the relationship of the efficacy with the values of the GoldScore scoring function and the values of total energy of ligand-receptor complex. This relationship was modeled with third degree of polynomial in Matlab. The GoldScore scoring function is used for the prediction of ligand binding positions and it takes into account factors such as H-bonding energy, van der Waals energy, metal interaction and ligand torsion strain. In contrast to it the Chemscore scoring function incorporates a protein-ligand atom clash term and an internal energy term. It takes account of hydrophobic-hydrophobic contact area, hydrogen bonding, ligand flexibility and metal interaction. Therefore, the aim of presented work is to find an optimal fitting polynomial function by which to model the relationship between quantitative parameters of {\it in vitro} bioassay (efficacy, affinity and potency) and the values obtained from molecular docking with crystal structure of DOR (PDBid:4ej4). The finding, established in this study, suggests that the third degree polynomial could be successfully used for modeling of the relationship between the efficacy of delta-selective enkephalin analogues and the results from the docking experiments. It is described by a polynomial surface of the third degree. This function could serve to predict the biological activity of new analogues and to be very useful in the design of new delta-selective analogues. Acknowledgments: This work is partially supported by the project of the Bulgarian National Science Fund, entitled: “Bioinformatics research: protein folding, docking and prediction of biological activity”, code NSF I02/16, 12.12.14.