Please note that the PDB file used for generating the molecular surface should not include hydrogen atoms. NMR structures will include hydrogens; delete the hydrogens from a copy of each structure and use that copy in MS.

Creating the Molecular Surface

What about hydrogens in molecular surfaces?

If you use the QCPE version of MS, you must run reformatms to convert the surface to the format used by sphgen (both formats are described in the reference manual section on reformatms). reformatms is interactive and requires the surface and the PDB file used to generate the surface.

How do I get QCPE MS to recognize standard PDB files?

Users of the UCSF MidasPlus package may use the output from the dms program directly as input for sphgen.

sphgen

Do I need to use sphgen?

Creating INSPH

Running sphgen

Looking at the Output

Getting a Good Sphere Cluster

The easiest way to fix a sphere cluster is to use graphics to identify spheres that you don't really need, then remove them. When you've found the unnecessary ones, go back to the original sphere cluster file (i.e. the one from sphgen) and delete the corresponding lines - the residue number in the PDB-like file of centers is the first number in the line in the sphere file. Remember to change the number of spheres listed on the line with the cluster number to reflect the deletions.

If your cluster is large - more than about 100 spheres - and deleting spheres by hand looks too tedious, you can use cluster to break it into smaller clusters. cluster is described in the reference manual; read the documentation completely before you try it. Start with the parameters given and experiment with the values; small changes can make a big difference in the result. Be aware that if the best cluster found is the same as the original input cluster, the program will appear not to have done anything.

The two methods just described may be combined if the best cluster output is not quite right. More spheres can be deleted from the new cluster, or, if the new cluster is too small, additional spheres may be chosen graphically. A cluster containing all the desired spheres may then be created by editing the sphgen output.

If nothing else works, it is possible to run cluster on all possible spheres rather than a preselected group. Use the analytical clustering algorithm in cluster on cluster 0, and experiment until you get what you want. Flagging spheres in important regions of the site may help.

Contact Scoring

distmap

distmap produces a score for each point in a cubic lattice based on how many receptor atoms are positioned to make favorable or unfavorable contacts with that point. The resulting grid will be used by DOCK to score each atom in a ligand orientation; the score of the orientation is the sum of the atom scores.

Positioning the Grid

distmap uses a PDB-format file as input; by default it will make a lattice which encloses all the atoms in that file. However, it is really only necessary to enclose the region of space where the ligand atoms will lie, plus the space containing those receptor atoms which might come close to ligand atoms. In other words, the grid should contain the site of interest and a generous amount of its surroundings but need not take in the whole protein.

The interactive program showbox is recommended for defining the size and shape of the grid. One way to do this is to make a box which encloses your sphere cluster and add an extra margin which encloses all the receptor atoms which might contact the ligands. The box generated is in PDB format; it should be viewed along with the receptor and possibly regenerated until it seems appropriate.

Creating INDIST and Running distmap

Input to distmap is placed in a file called INDIST, which is described in the reference manual. You will need to change pdbnam and scoren, the input and output file names, but the rest of the parameters given in the example are reasonable starting values. Be sure to include the name of the box file from showbox on the last line. distmap, like sphgen and all the programs that use similar input files, must be started while in the directory where INDIST is located.

The Electrostatic Potential Map (Delphi)

DelPhi may be used to calculate the electrostatic potential due to the receptor. DOCK can use this potential grid to calculate electrostatic scores for ligands with partial charges.

DelPhi is not supplied with DOCK; it is available from Barry Honig, Columbia University, or from Biosym. If you have the program and wish to use it for electrostatic scoring, follow the directions supplied with it to generate a potential map for your protein. DOCK currently only reads the potential map (phi file) format used by version 3 of DelPhi.

What's the scoop on Delphi scoring with DOCK?

Force-field scoring

chemgrid

chemgrid saves information about the steric and electrostatic environment at each point on a grid, so that ligand orientations can be scored rapidly during a DOCK run.

Positioning the Grid

You determine the location and dimensions of the region to be gridded by using the program showbox to create a box which contains the desired region. For chemgrid, the box should enclose the volume that the ligand orientations are likely to occupy (note that this is slightly smaller than the volume required for distmap). An easy way to accomplish this is to generate a box which encloses the spheres to be used for docking along with an extra margin. The box generated should be viewed along with the receptor and possibly regenerated until it looks good. Unlike the limiting box that can be used as input to distmap, the grid region does not determine which receptor atoms are used in the calculation; all receptor atoms within the specified cutoff distance of any grid point will be included.

Preparing the Receptor File

In order for the steric and electrostatic environment within the site to be evaluated, the program will need to associate each of the receptor atoms with the proper charge and steric qualities. This is accomplished by matching their names to names in a parameter file. Three parameter files for macromolecules are supplied with this release:

na.table.ambcrgna.table.ambcrg    nucleic acid parameters
prot.table.ambcrg.ambHprot.table  protein parameters; Amber hydrogen names
prot.table.ambcrg.pdbHprot.table  protein parameters; PDB hydrogen names

All three contain AMBER united-atom parameters. The only explicit hydrogens included are those bonded to polar atoms (anything except carbon). The protein files differ only in the hydrogen-naming conventions used; the heavy atom names are PDB standard in both. In your receptor file, the atom names should match the names in the parameter file you will be using, and hydrogens bonded to polar atoms should be present. It is all right to have all the hydrogens and even lone pairs present since any atom not found in the parameter file receives zero charge and volume. Special attention should be given to the names of atoms at the termini and the residue names for histidine and cysteine. The different protonation states of histidines correspond to different residue names: HIP for positively charged (hydrogens on both nitrogens), HID for neutral with the delta nitrogen protonated, and HIE for neutral with the epsilon nitrogen protonated. CYS refers to a cysteine with a free sulfhydryl group; CYX refers to a cysteine involved in a disulfide bond (a half-cystine). Note that some structures in the PDB use CYS in disulfides; these should be edited to CYX.

Creating INCHEM and Running chemgrid

Input to chemgrid is placed in a file called INCHEM, which is described in the reference manual. The input values should be placed in the file as shown in the example. You will need to replace recfil with the name of your receptor file. Replace table and vdwfil with the locations of your chosen parameter table and vdw.parms.amb on your system. Inbox should be replaced with the name of the showbox output file.

Grddiv values between 0.2 and 0.5 are recommended; fine grids are preferred. Any combination of grid point spacing and x, y, and z dimensions can be used as long as the number of points does not exceed the maximum array size specified in the program. If this happens, you will get a message when you run chemgrid. To resolve the problem, the grid spacing may be increased or the box dimensions may be decreased. DOCK may also be recompiled with larger array sizes if your computer's memory allows; edit the file chemgrid.h appropriately.

A dielectric function of 4.0r or 4.5r and a cutoff of 10.0 ngstroms or more are appropriate in most cases. (This dielectric corresponds to an estype of 1 and esfact of 4.0 or 4.5.) If a constant dielectric is selected, an "infinite" cutoff (one large enough to include the whole receptor) should be used. We tend to use close contact limits of 2.0-2.5 and 2.5-3.0 ngstroms for receptor polar and nonpolar atoms, respectively. The close contact limits do not affect the force field scores that orientations receive, but they determine which orientations are thrown out when the force field scoring only option is used in DOCK. The resolution of the protein structure should be kept in mind when setting these limits; if the receptor atom positions are not very well-defined, it is best not to constrain the results too strongly based on these positions.

Grdfil will be the beginning of the output file names.

Although we have recommended certain values, we would like you to try whatever you feel is appropriate, based on your knowledge of the parameters and their significance.

Parameterization takes place in the first few seconds of a chemgrid run, and produces two files: PDBPARM, which lists the coordinates of each receptor atom together with the associated parameters, and OUTPARM, which reports each atom not found in the parameter file and the apparent net charge of the receptor. A long list of hydrogen atoms is normal, since there are no parameters for hydrogens not attached to polar atoms. It is important to check the net charge since a strange value can alert you to parameterization problems. You may use the script called charge, supplied with DOCK, to estimate the expected charge. A value which differs from that expected may mean that something is wrong with the hydrogens or the residue names. Another common cause of a non-integral net charge is incompletely modeled residues.

Three output files, named grdfil.bmp, grdfil.esp, and grdfil.vdw, make up the actual grid.

What's all this stuff in my OUTPARM file?

mol2db is also interactive; it reads SYBYL ASCII (MOL2) format. If you wish to use contact scoring only and your ligands have no charges, choose version 2.0 output. Otherwise, your ligands should have hydrogens and charges and you should create a version 3.5 database. mol2db may be used to label ligand atoms for chemical matching, described below.

DOCK databases may include any number of molecules. They consist of lists of molecule records; you may edit them after they are created to delete molecules or to separate some molecules into smaller databases. If you create smaller files from a DOCK 3.5 database, be sure to include the header information in each file. (The first line should read DOCK 3.5 ligand_atoms, and the chemical matching information should follow it starting on the second line. You can copy the header from the beginning of the original database.)

Labeling Atoms and Spheres for Chemical Matching

colsph may be used to label spheres according to the electrostatic potential at their location in space. You will need a sphere cluster file and either a potential map from DelPhi or the force field grid files created by chemgrid. Create a file specifying the labels and the range of receptor potential or of the electrostatic potential component of the force field grid which corresponds to each label. On each line, list one label and the upper and lower bounds of its electrostatic potential range. The potential will be evaluated at each sphere and the sphere will be assigned a label if the potential is in the appropriate range. Run colsph interactively. The program will prompt you for the map type and name, the name of the range file you just created, and the input and output sphere file names. (colsph will color spheres in all clusters in the sphere file.)

To color ligand atoms, begin with a list of molecules in SYBYL ASCII (MOL2) format. Run mol2db interactively (see the reference manual). At the coloring prompt, enter one label at a time with its corresponding SYBYL atom type. To indicate that a given atom type should be (or not be) in a particular functional group, you may include on the same line a second atom type and the number of bonds which separate it from the first atom. Atoms of the first type will then be labeled only if they are the specified number of bonds from an atom of the second type. If the number of bonds is negative, atoms of the first type which are the specified number of bonds from the second type will not be labeled (this is useful for excluding some atoms which meet a previous labeling criterion). Enter a blank line to end label entry.

In your INDOCK file, you will need to specify which of the labels you have assigned to spheres should match which of the ligand atom labels.

Tell me more about chemical matching...