The purpose of this tutorial is to become familiar with how to create a complex of two molecules via DOCKING the two molecules together. There are at least three methods used to dock two molecules. In this tutorial, we will attempt the first two methods. The third method is too computationally expensive to attempt during the tutorial, but you are encouraged to use this technique during your Individual Modeling Project activities.

Docking with a grid was very popular during the 1980s when computers were not very powerful. Even though computing power has improved during the last decade, the grid method is still fairly popular, unless both molecules are relatively small.

A variation of this method apply a step-wise approach to the distance restraint's force constant:

• a low-force-constant (~5-10 kcal/mol) distance restraint is applied during the early stages of energy minimization until the non-fixed molecule is near the surface of the fixed molecule
• then the force constant is set to a high value (~100 kcal/mol) and the non-fixed molecule is minimized again so that the putative interaction is guaranteed to be present in the final structure. This final step also allows the modeler to quickly determine if multiple docking calculations were all successful just be observing the energy---if one calculation has a significantly higher energy, this usually implies that the putative atomic interaction was not properly formed, and the final minimization is trying to "bend" and "torque" the non-fixed molecule to satisfy this large energy penalty.

Another variation applies a high-force-constant (~100 kcal/mol) distance restraint so that the distance is only 2-3 angstroms shorter than the current distance. This minimization is repeated again and again; the distance between the atoms is restrained to a value 2-3 angstroms shorter than in the previous minimization for each round of minimization.

Other variations apply simulated annealing to allow the non-fixed molecule to avoid high-energy minima, which (presumably) culminate in "poor" molecular complexes.

AutoDock

1. The temperature of the system is set to a high value (~500K)
2. The coordinates of the ligand are changed randomly
3. the energy of interaction between the two molecules is calculated
4. The Monte Carlo Method is applied:
   1. If the energy of interaction is lower than the previous step, this new conformation is ACCEPTED.
   2. Iif the energy of interaction is higher than the previous step, the difference in energy is calculated; the probability (between 0 and 1) that the energy could increase by this amount is then calculated based upon the Boltzman distribution at the system's temperature. A random number (between 0 and 1) is chosen. If the random number is larger than the probability, then this new conformation is REJECTED; if the random number is smaller than the probability, then this new conformation is ACCEPTED. Therefore, some conformations that are HIGHER in energy than the previous step can be accepted.
5. Step 1 is repeated until the number of ACCEPTED or REJECTED steps reaches a pre-set limit (e.g., 100,000 steps)
6. The temperature is reduced by a few percent, and step 1 is repeated at this new temperature. Note that the number of accepted steps that are higher in energy than the previous step will be high when the temperature is high, and will be lower as the temperature is lowered.
7. Once the final temperature reaches a very low value (~1K), the Monte Carlo simulation is stopped.

Part 1: Preparing the System

1. Download the cruzain.psv file.
2. Start InsightII and restore the cruzain.psv folder.
3. Fix/Fix/Fix your potentials/partial charges/formal charges.
4. Measure bumps and hydrogen bonds between the ligand and the protein with MONITOR ON

This complex consists of the protein cruzain and a ligand (an inhibitor); this system was kindly provided by Karl Scheidt, a member of William Roush's research group at the University of Michigan. All atoms of the inhibitor, and all atoms of residue 25 in cruzain are shown, and are colored by atom type. Except for residue 25, only the trace (the alpha-carbons) of cruzain is shown (in purple). This initial visualization was chosen for it's information content: to show how the ligand fits rather well in the active site. Is this visualization a SIMPLIFICATION, DIDACTIC ILLUSTRATION, ANALOGY, or NUMERICAL REPRESENTATION? To view hydrogen bonds, VDW overlaps, etc., you may have to visualize (change the display to gain more information, e.g., display all atoms).

There are two hydrogen bonds formed between residues 66 & 158 (in white) and the ligand:

Atom 1	Atom 2	Distance	Dihedral Angle
CRUZAIN:66:HN	LIGAND:I2:O	2.32	153.51
LIGAND:I3H:HN	CRUZAIN:158:O	2.13	141.77

Which hydrogen bond is "stronger"? Why?

There are many van der Waals bumps; most of these involve residue 25 (residue 25 is colored by atom type):

Atom 1	Atom 2	Distance
CRUZAIN:25:SG	LIGAND:I3H:CA	2.72
CRUZAIN:25:CB	LIGAND:1C:H12	2.08
CRUZAIN:25:HB2	LIGAND:1C:H22	1.89
CRUZAIN:25:SG	LIGAND:1C:C1	1.89
CRUZAIN:25:SG	LIGAND:1C:H12	0.85
CRUZAIN:25:SG	LIGAND:1C:H11	2.44
CRUZAIN:25:SG	LIGAND:1C:C2	2.96
CRUZAIN:25:HG	LIGAND:1C:C1	2.05
CRUZAIN:25:HG	LIGAND:1C:H12	1.55
CRUZAIN:25:HG	LIGAND:1C:H11	1.91
CRUZAIN:67:HD21	LIGAND:I2:HN	1.78
CRUZAIN:158:O	LIGAND:I3H:HN	2.13

This complex is actually a modification of the crystal structure: Originally, the crystal structure had a covelent bond between the sulfur of Cysteine-25 and the C1 carbon of the ligand. To prove this, measure the distance between these two atoms; keep "monitor" ON. Is the a typical distance of a C-S bond? Remember this value---you will need it later! This bond has been removed and hydrogens have been added to the sulfur and carbon, causing several large VDW overlaps.

In this assignment, we will dissociate the ligand and the protein, and then attempt to model how the ligand binds to the protein BEFORE the covalent bond is created. One of the biggest problems with this exercise is that the protein will NOT be allowed to move. Therefore, the protein is locked into it's conformation WITH the covelent bond, while we are assuming that the covalent bond is NOT yet made.

1. Click on the Connect Object button and connect to the ligand. Move the ligand until the distance between the C1 carbon atom of the ligand and the SG sulfur atom of cruzain-25 is ~15 angstroms (this distance is already monitored and updated on the screen if you followed the instructions in the discussion above).
   
2. Docking while displaying hydrogen bonds and bumps can be slow. Turn OFF the monitored hydrogen bonds and bumps!

Part 2: Manually Docking with a Grid

1. Select the Docking Module.
2. Click on Docking_Grid/Create, and create an enclosure-style grid about cruzain, with a border space of 5 angstroms and a 2 angstrom grid step. This grid has "coarse resolution"---fewer grid points means lower accuracy, but few grid points to evaluate means faster calculation times per docking step. For your research, you should use a grid with finer resolution. However, for the purposes of this tutorial, we will sacrifice accuracy for speed and use a grid with coarse resolution.
3. Click on Docking_Grid/Compute, and compute a docking grid of cruzain with Van_der_waals and Coulomb energies and a cutoff of 10 angstroms. This can take several minutes to compute. Mmake a visible grid by turning on make_vis_grid with a grid name of CRUZAIN_VGRD.
4. Click on Grid/Display, and display all points of CRUZAIN_VGRD.
5. Click on Grid/color and color all grid points using the charge_spectrum (turn "use_spectrum" ON)
6. Click on Grid/Display, and turn off the display of the grid points
7. Evaluate grid slices:
   Click on Grid/Slice
   Set the Scalar grid name to CRUZAIN_VGRD
   Set the Plane Number to 1
   Set the Plane Direction to Z
   Set the Plane Height to 59.5221
   Set the H intvals to 20
   Set the Plane Width to 59.5221
   Set the W intvals to 20
   Set the Plane Type to Mapplane
   Set the Spectrum Name to GRD_SPECTRUM or CHARGE_SPECTRUM
   Set the Plane Style to Filled
   Click execute. Move the plane by moving the slide bar in the Parameters menu. Notice that when the plane bisects the molecule, the center of the plane is blue: the interaction energy inside the molecule is high (positive, or BAD). When the plane is moved towards the ligand and outside the protein, the plane turns red: the interaction energy is low (negative, or GOOD). Position the plane near the ligand; notice that the plane is red in the center, and blue near the edges: the interaction energy is GOOD in the center and BAD at the edges, so the ligand is attracted towards the center---the active site! This model may actually be useful!

   Try a Contour plane in the X direction with a SLICE_SPECTRUM, Contour_Levels set to -2, and Displacement set to -6
   Add a Plane Number 2 at a contour level of -3.
   Add a Plane Number 3 at a contour level of -4.
   Add a Plane Number 4 at a contour level of -5.
   Add a Plane Number 5 at a contour level of -6.
   Add a Plane Number 6 at a contour level of -7.
   Notice how the interaction energy becomes lower (more negative) nearer the active site.

   Try other planes and parameters. When you are done, set Plane Type to OFF for all planes.

8. Evaluate Grid Contours
   Click on Grid/Contour
   Set Contour_Name_Root to contour1
   Set Level Specification to Single
   Set Contour Level to -15
   Set Display Style to Solid
   Set Color to yellow
   Note where the best interaction energies lie. Try other parameters (e.g., Contour Level = -10, -7, -5). Then delete all contour* objects.
9. Evaluate the Intermolecular Energies
   Click on Evaluate/Intermolecular, and ADD a MONITORed energy between CRUZAIN and LIGAND via the GRID.
   The VdW, Elect, and Total energies are reported on the screen. Connect to the ligand and move it towards the active site. Note that the energies are updated fairly rapidly. Place the ligand in or near the active site. This may take ~5-10 minutes. DON'T spend more than 10 minutes on this step. Be sure to turn OFF any contours or slices so that the graphics don't make this slow. When you are comfortable with this display, turn ON intermolecular bumps.
10. Position the ligand so that the VdW and Elect energies are < 0, and the Total energy is < -10.
11. Position your system for best visualization of the ligand-receptor interaction. Annotate your model with text and an arrow. Choose any text and any arrow location.
12. Save a folder of your model using the name manual_dock.psv.
13. Clear all bump monitors.
14. Click on Evaluate/Intermolecular, and ADD a MONITORed energy between CRUZAIN and LIGAND. DO NOT use the grid. Use a cut-off of 10 angstroms, with both Van_der_waals and Coulomb energies. Move the ligand around. Notice that the calculation takes a LOT longer! Don't spend more than a minute using this mode.
15. Delete all objects.

Part 3: Traction Beam Docking

1. Read the cruzain.psv folder into InsightII.
2. Fix/Fix/Fix your potentials/partial charges/formal charges.
3. Connect to the ligand. Move the ligand until the distance between the C1 carbon atom of the ligand and the SG sulfur atom of cruzain-25 is 15 angstroms (monitor this distance).
4. Select the Assembly/Associate menu and associate cruzain and the ligand using a New Assembly Name of "complex".
5. Select the Discover module.
6. Select the Restraints/Fix menu and fix the entire cruzain molecule.
7. Select the Restraints/Generic Dis menu and set a generic distance constraint between CRUZAIN:25:SG and LIGAND:1C:C1
   Set GenericDis_Upr_Bnd to 2.73
   Set GenericDis_Lwr_Bnd to 2.73
   Set GenericDis_Upr_K to 5.00
   Set GenericDis_Lwr_K to 5.00
   Set GenericDis_Max_Frc to 15
8. Minimize using Conjugate Gradients (the structures are already at a fairly low minimum energy, so we don't have to use steepest decents) to a derivative of 0.5 or 1500 iterations, whichever comes first. Use charges, but don't use cross or morse terms. Set a dielectric of 1.0, with no distance dependence (use the Discover/Parameters/Set menu). Set a cutoff of 20 angstroms (use the Discover/Parameters/Variables menu). Select the Run/run menu and INTERACTIVEly RUN_MINIMIZATION on the COMPLEX.
9. If the ligand is not within 5 angstroms of the protein, repeat the previous step.
10. Change the restraints so that the GenericDis_Upr_K = 50.00, the GenericDis_Lwr_K = 50.00, and the GenericDis_Max_Frc = 150.
11. Repeat the minimization with the new distance constraint. Choose an appropriate number of iterations and derivative so that the final complex is a "minimum". Of course, this depends on YOUR interpretation of a "minimum"!
12. Continue to repeat until the "minimized" ligand is within 3.15 angstroms of the protein.
13. Save the final result in a folder named traction_dock.psv in your home directory.

Part 4: Verify that you have completed this portion of the assignment

See the Docking/Ligand Design/Electrostatics Assignment page for details.