Page 264 - Big Data Analytics for Intelligent Healthcare Management
P. 264

10.7 MOLECULAR INTERACTIONS AND SCORING FUNCTIONS               257





                            Protein                MGL tool               Ligand
                            preparation                                   preparation




                        File—read molecule—select protein      Ligand—input—open the ligand





                        Edit—hydrogen bond—add—polar only  Ligand—torsion tree—choose torsion—done



                        Grid—macro molecules—choose—protein—    Ligand—output—save as pdbqt
                        save as pdbqt




                            Grid—grid box—set grid box

               FIG. 10.2
               Workflow steps for preparation of proteins and ligands in AutoDock


               The genetic algorithm is applied to bimolecular systems to find the closest conformation with minimum
               global energy. It is based on the genetic operator. Compared to other algorithms, in DG algorithms a
               smaller set of distance constraints is used [51]. The genetic algorithm is used to find approximate so-
               lutions for search problems of global minimum energy. It is based on the genetic operator, which com-
               bines two chromosomes to produce new chromosomes. In this process, many complex scoring
               functions are used. Lamarckian genetic algorithm (LGA) is also used in docking algorithms, which
               controls genotype space and phenotypic space. The former occurs in mutations while phenotypic space
               is determined by the energy function optimized. In energy minimization, the local sampling is per-
               formed after the genotypic changes [52]. The workflow steps for preparation of proteins and ligands
               and the framework for preparation of configuration files for running through AutoDock vina, are shown
               in Figs. 10.2 and 10.3 respectively.





               10.7 MOLECULAR INTERACTIONS, SCORING FUNCTIONS, AND DISCUSSION
               OF SOME DOCKING EXAMPLES
               Docking algorithms anticipate many orientations of the ligand in the binding pocket of the receptor. In
               docking simulations, the process shows small molecules based on the database. Docking results are
               ranked depending on the degree of binding energy between the ligand and receptor. The correctness
               of docking depends on the quality of scoring function [53]. Scoring functions are based on mathemat-
               ical approximation methods for assessment of the binding energy. Docking at the molecular level is a
   259   260   261   262   263   264   265   266   267   268   269