Protein Docking and Interactions Modeling CS374 Fall 2004 Lecture 12, 11/04/04Lecturer: Maria Teresa Gil Lucientes Scribe: William LiuProtein Docking and Interactions ModelingBased on the following papers:1. Gray J., Moughon S., Wang C., Schueler-Furman O., Kuhlman B., Rohl C., Baker D., “Protein-Protein Docking with Simultaneous Optimization of Rigid-body Displacement and Side-chain Conformations”, J. Mol. Biol., 333:281-299, 2003Additional References:1. Halperin I., Ma B., Wolfson H., Nussinov R. “Principles of Docking: An Overview of Search Algorithms and a Guide to Scoring Functions”, Proteins: Structure, Function, and Genetics, 47:409-443,20021 MotivationThe lecture presented the basic goals of protein docking: whether two molecules “interact” and whydocking is important. Namely, protein dockings relevance to cellular biology as well as its importanceto pharmacology and rational drug design. The major types of docking: protein-protein docking andprotein-ligand docking were discussed as well as the major techniques of docking: surfacerepresentation and surface matching. These discussions lead to a particular protein-protein dockingalgorithm that uses low-resolution Monte Carlo search and Monte Carlo minimization. In closing,there were some conclusions presented about the complexity of docking and the obstacles which needto be further explored.2 Basics2.1 Goals and Significance of Protein DockingThe fundamental goal of protein docking is to determine whether two molecules interact and if so, theorientation that maximizes this interaction as well as minimizing the total energy of the resultingcomplex. Ultimately, we would like to have a database of molecular structures that would yieldinformation on all other molecules that could interact with the given structure. 1Protein Docking and Interactions Modeling CS374 Fall 2004 Lecture 12, 11/04/04Lecturer: Maria Teresa Gil Lucientes Scribe: William LiuFigure 1: HIV-1 ProteaseFigure 2: HIV-1 Protease with inhibitorDocking is imminently important in many areas. Specifically, in cellular biology, the function ofproteins is a result of its interaction (i.e. docking) with other proteins as well as other molecularcomponents. Therefore if we could predict how/if proteins interact (dock) with other molecules wecould possibly infer or inhibit function. The latter statement is of particular interest to drug andpharmaceutical companies. Thus the results of docking can extremely beneficial in finding drugs thatare effective against particular diseases. A concrete example can be seen with HIV-1 Protease:Since the active site (Aspartyl groups) for HIV-1 Protease (Figure 1) is known, constructing a ligand tobind with this region would render the HIV-1 Protease ineffective in replication (this class of drugs isknown as protease inhibitors). Thus, if we could solve the docking problem, finding a drug that wouldbe effective for this particular disease would simply require iterating through our large library ofligands and seeing which ones dock with this particular receptor site.However, predicting whether two molecules bind is a complex problem. Since we are dealing withtwo structures which are 3 dimensional, the 6 degrees of freedom (translations and rotations in the x, yand z axes) already pose a complex search space. Given that these structures are not completely rigid,especially when binding, further complicates our problem. This flexibility of the molecules results inhundreds to thousands of degrees of freedom and an astronomical number of conformations2.2 Types of DockingProtein-Protein DockingIn these docking situations, the protein molecules are usually considered rigid bodies. Thus, we canapproximate the problem with 6 degrees of freedom: the 3 possible translations and rotations in the x,y and z axes. To further reduce the search space of the problem, steric constraints (i.e. no overlap inVan der Waal envelope, geometric constraints) are applied. Then energies of the resulting bindingconfirmations are examined. In essence this type of docking consists of constructing the 6dimensional confirmation space of all geometrically/physically allowable configurations, then findingthe best configuration by comparing energy scores of the resulting binding complexes. Protein-Ligand DockingThis type of docking is much more complex because of the flexibility of the ligand. The ligandintroduces this complexity because of rotatable bonds (Figure 3) that exist within the molecule. As thename suggests, these bonds allow parts of the molecule to rotate, thus increasing the dimensionality ofthe problem with each rotatable bond. 2Active SiteProtein Docking and Interactions Modeling CS374 Fall 2004 Lecture 12, 11/04/04Lecturer: Maria Teresa Gil Lucientes Scribe: William LiuFigure 3To model the complex protein-ligand docking, it is usually common to either reduce the flexibility ofthe ligand (i.e. consider it rigid, or consider only a specific number of rotatable bonds) or search theconformational space using either Monte Carlo methods or molecular dynamics.2.3 Docking TechniquesSurface Representation This technique is used to represents the docking surfaces of the protein and identifies possibleregions of interaction (i.e. cavities and protrusions). Connolly Surface:To model the surface of the molecules, the atoms which are accessible to a probe sphere(atoms making up the contact surface) are represented with their Van der Waal's surface.Thus, we piece together these external patches of convex, concave, and saddle-shapes to getthe resulting Connolly Surface (Figure 4).Lenhoff:Instead of computing the Connelly surface, this technique computes the possible locations ofthe ligand that will be bound. Thus a “complementary” surface is computed for the receptor. Figure 4: Convex regions are yellow, concave regionsare blue and saddle-shaped pieces are greenFigure 5: White dots represent possible location of ligand atom centersKuntz et al. Clustered Spheres:Generate a sphere for every pair of points i and j that lie on the surface. The generated sphereis centered on the normal (perpendicular) at point i. Therefore, regions where these generated3Protein Docking and Interactions Modeling CS374 Fall 2004 Lecture 12, 11/04/04Lecturer: Maria Teresa Gil Lucientes Scribe: William Liuspheres overlap define possible areas of cavities on the receptor and
View Full Document
Unlocking...