Kuntz Home / DOCK Home / DOCKumentation Contents / Beginner's Guide

Introductory Remarks

Scope Up / What DOCK Can Do For You / Basic References / Overview / Disk Space Caution

Scope of this Guide

This section is intended as a supplement to the DOCK reference manual. It describes the steps a new user would typically take to apply the programs to a macromolecule and potential ligands of interest. While the reference manual describes in detail the various input and output files, this guide is meant to convey in informal terms the process as a whole. Some of the difficulties we have encountered as well as approaches we have found useful are discussed.

What DOCK Can Do For You

DOCK is a program for locating feasible binding orientations, given the structures of a "ligand" molecule and a "receptor" molecule. What is considered feasible depends on how the orientations are evaluated. Current options are a contact (shape-fitting) score, a force field interaction energy, and an electrostatic energy calculated by using a DelPhi potential map (the program DelPhi is not distributed with DOCK). Atoms may be labeled so that they may fall only in chemically appropriate regions (as labeled by the user), and orientations may be varied to optimize their force field scores. In SINGLE mode, DOCK generates many orientations of one ligand. In SEARCH mode, orientations are generated for each of the molecules in a database in turn; the best-scoring orientation of each molecule is saved, and the best-scoring molecules are written out. Some of the molecules in the list of best-scoring compounds, perhaps with modifications, may be interesting as potential new ligands for the receptor.

Basic DOCK References

These reviews give a general picture of DOCK and an overview of its applications:

Kuntz, I.D. (1992) Science. 257: 1078.
Kuntz, I.D., Meng, E.C., and Shoichet, B.K. (1994) Acc. Chem. Res. 27: 117.

New users should become familiar with the algorithms used by DOCK. Reading these papers is strongly recommended:

Kuntz, I.D., Blaney, J.M., Oatley, S.J., Langridge, R., Ferrin, T.E. (1982) J. Mol. Biol. 161: 269. The original DOCK publication. This paper outlines the docking process and describes the way spheres are generated. Corresponds to DOCK version 1.0.
Shoichet, B.K., Bodian, D., and Kuntz, I.D. (1992) J. Comp. Chem. 13: 380. The current methods for generating matches between ligand atoms and receptor spheres, introduced in DOCK 2.0, are described here.
Meng, E.C., Shoichet, B.K., and Kuntz, I.D. (1992) J. Comp. Chem. 13: 505. Describes the use of DelPhi electrostatic potentials and force-field-like electrostatic and van der Waals potentials for scoring. DOCK 3.0.

These papers are also worthwhile reading, especially if you plan to use the features and techniques they describe:

DesJarlais, R.L., Sheridan, R.P., Seibel, G.L., Dixon, J.S., Kuntz, I.D., and Venkataraghavan, R. (1988) J. Med. Chem. 31: 722. The first use of DOCK to search molecule databases. DOCK version 1.1.
Shoichet, B.K., and Kuntz, I.D. (1991) J. Mol. Biol. 221: 327. Discusses protein-protein docking.
Meng, E.C., Gschwend, D.A., Blaney, J.M., and Kuntz, I.D. (1993) Proteins: Structure, Function, and Genetics. 17: 266. Introduces the use of minimization to improve the scores of promising matches from DOCK. Minimization is a new option in DOCK 3.5.
Shoichet, B.K., and Kuntz, I.D. (1993) Protein Engineering. 6: 223. Introduces the concept of labeled or colored matching (the implementation in DOCK 3.5 is slightly different from what is described in this paper).

An extended list of Kuntz group papers referring to DOCK is in the References section of the reference manual.

Overview of the DOCK Package

The basic requirement for docking is a structure of the macromolecule of interest. The docking procedure can be divided into four general stages: site characterization, calculation of grids for scoring, preparation of databases, and DOCK itself.

For DOCK, sites on the receptor are characterized by sphere clusters. These are simple geometric descriptions of the volume available to ligands. The program sphgen calculates these sphere clusters using the molecular surface of the receptor produced by Connolly's MS program. (MS should be obtained from the Quantum Chemistry Program Exchange).

While site characterization prepares the information needed to generate ligand orientations, grid calculations are necessary so that the orientations can be evaluated, or scored. The kind of scoring desired dictates which programs need to be run. For contact scoring, distmap is used to generate the grid. This grid is also necessary for DelPhi electrostatic scoring, since it determines which orientations must be thrown out due to bad contacts with receptor atoms. Electrostatic scoring requires a potential map from DelPhi (Honig et al., Columbia University). The grid used for force field scoring, which includes both steric and electrostatic terms, is produced by chemgrid. The chemgrid result may be used alone or combined with the distmap grid to use both contact and force field scoring.

If DOCK will be used to search a database of potential ligands, their structures must be converted to DOCK 3.5 database format; mol2db can create the database from an input list of ligands in SYBYL ASCII (MOL2) format. mol2db can also be used to label atoms by type and chemical environment so that they can be restricted to matching only chemically appropriate spheres. The spheres are given labels using colsph.

The final stage of the process is running DOCK and viewing the results. DOCK attempts to match ligand centers to receptor sphere centers, then scores each orientation using the information in the precalculated grids. At the user's option, DOCK can attempt to adjust ligand orientations to optimize their scores. The best-scoring molecules or orientations may be viewed using a molecular graphics program.

A necessary side effect of having many options that the user can control is requiring the user to enter many numbers, which can result in confusion. We hope to make the learning curve an easier place to be with this guide.

A Caution Concerning Disk Space

The output from some of the programs associated with DOCK, particularly MS, sphgen, and DOCK itself, may require substantial amounts of disk storage. It is a good idea to be cautious at first; check before starting your job to make sure there is space available. While DOCK jobs are running, check to be sure they are not creating overly large files, especially if you have increased the bin parameters.

Curator: Daniel Gschwend, gschwend@cgl.ucsf.edu (rev. 1 September 1995)