« CoMFA practical/tutorial
Free data mining software »

Docking practical/tutorial

Other than the CoMFA practical, I have also prepared a docking practical. All the files needed for the practical (except Sybyl 7.2) can be downloaded from here.

In this practical, we will be looking at two different methods for ligand-protein docking: DOCK and FlexiDock.

Protein

  1. Load a protein (HIV-1 protease).
    • File >>> Read.
    • Select protein.pdband press OK.
    • Select YES in the Center the molecule dialog and press OK.
  2. Name the protein.
    • Build/Edit >>> Name Molecule.
    • Type HIV_protease for the new molecule name (press OK).
  3. Add hydrogens to the protein.
    • Biopolymer >>> Prepare Structure >>> Add Hydrogens.
    • Press OK.
  4. Compute charges for this protein.
    • Biopolymer >>> Prepare Structure >>> Load Charges.
    • Press OK.
    • Press Yes.
  5. Save the protein
    • File >>> Save As.
    • Type hiv_protease for the new molecule name (press Save).
  6. Zoom out to view the entire protein.
  7. Find secondary structures in protein.
    • Biopolymer >>> Conformation >>> Find Secondary Structure.
    • Use the Kabsch-Sander method.
    • Select Render Conformations and press Find.
    • Press Close.
  8. Construct a three-dimensional shaded ribbon for the protein.
    • View >>> Biopolymer Display >>> Shaded Ribbon.
    • Press All and press OK.
    • Select WHITE as the color and press OK
  9. Highlight catalytic triad ASP25, THR26 and GLY27 by spacefill representation.
    • View >>> Mixed Rendering.
    • Press Substructures and select ASP25, THR26 and GLY27 from both chain A and B and press OK.
    • Press OK.
    • Select SpaceFill for Representation and press OK.
  10. Rotate, translate, zoom in and out to have a good overview of the secondary and tertiary structure of the protein.
    • Question 1: Where is the active site of the protein?
  11. Delete the protein.
    • Build/Edit >>> Zap (Delete) Molecule.

 

Ligand

  1. Load a drug (A77800).
    • File >>> Read.
    • Select ligand.pdb and press OK.
    • Select YES in the Center the molecule dialog and press OK.
  2. Name the ligand.
    • Build/Edit >>> Name Molecule.
    • Type A77800 for the new molecule name (press OK).
  3. Add hydrogens to this drug.
    • Build/Edit>>> Add >>> Hydrogens.
  4. Compute charges for this drug.
    • Compute >>> Charges >>> Gasteiger-Huckel.
    • Press No when asked if you want to change formal charges before computing charges.
  5. Conduct molecular mechanics computation to find lowest potential energy conformation for this drug.
    • Specify the force field and type of charges to use.
      • Compute >>> Minimize.
      • In the Minimize dialog, press Modify.
      • In the Energy dialog, select Tripos from the Force Field option menu.
      • Select Gasteiger-Huckel from the Charges option menu.
      • Press OK.
    • Set the minimization parameters.
      • In the Minimize dialog, press Minimize Details.
      • In the Minimize Details dialog, increase the maximum number of iterations (Max.Iterations) to 10000.
      • Decrease the Gradient to 0.005.
        The Gradient value is a termination criterion. If the gradient difference between two consecutive iterations goes below this value, the calculation stops.
      • Set the color option to Force.
        This color codes atoms according to the total force on each atom, as the minimization proceeds.
      • Press OK.
    • Submit the job to run interactively.
      • In the Minimize dialog, type A77800 as the job name.
      • Press OK.
        As the minimization proceeds, changes to the structure are interactively displayed. Information about each iteration is also printed to the textport.
      • Question 2: What is the total energy of the minimized structure?
  6. Save the drug
    • File >>> Save As.
    • Type A77800 for the new molecule name (press Save).
  7. Create molecular surface of this drug.
    • Create a molecular surface.
      • View >>> MOLCAD Surfaces >>> Molecular Surfaces
      • Press Create in the MOLCAD Surfaces dialog.
    • Select type of molecular surface.
      • In the Create MOLCAD Surface dialog, select Connolly from the Type menu.
      • Press OK.
      • Press Done.
  8. Rotate, translate, zoom in and out to have a good overview of the conformation of the ligand.
  9. Delete the drug.
    • Build/Edit >>> Zap (Delete) Molecule.

 

DOCK

SYBYL’s docking functionality provides a real time approximation of the intermolecular energy of interaction between a pair of molecules (in kcals/mol), a useful tool for interactively identifying possible binding conformations. One molecule (stationary, by convention) is called the site, and the other is the ligand. Interactive output includes the total energy, the magnitude and direction of the overall force (for ligand atoms strongly interacting with the site), and the one site atom which is interacting most strongly.

  1. Load protein (hiv_protease.mol2) and drug (A77800.mol2). It is recommended that protein be loaded first.
  2. Position the ligand into the active site of the protein (you may wish to turn on shaded ribbon for the protein and spacefill for ligand).
  3. Prepare DOCK parameters.
    • Options >>> Tailor.
    • In the Tailor dialog, select DOCK from the Subject menu.
    • Increase the maximum number of lattice points (MAX_LATT_PTS) to 800000.
    • Press Apply and then press Close.
  4. Begin docking.
    • Compute >>> Dock.
    • Select M2:A77800 as the ligand molecule area (press OK).
    • Select M1:HIV_protease as the site molecule area (press OK).
      Three new items appear on the screen:
      1. An Energy Gadget dialog, showing the total energy of interaction, plus its steric and electrostatic components, for the currently displayed configuration using the Tripos force field.
      2. A box surrounding the ligand molecule. As you move the ligand, the energy of interaction is computed only for atoms within that box.
      3. A prompt at the command line for the next action SYBYL is to take.
        • press ? to show other actions that can be taken from within the docking process.
        • If you wish to stop DOCK at any time, press to exit the DOCK command.
  5. Move the ligand.
    • Slowly move the ligand a few Å in any direction.
      As an atom of the ligand gets too close to a site atom, the energy of interaction suddenly increases, and new yellow and red lines are displayed.
      • Yellow lines connect ligand atoms to the nearest bumping site atom,
      • Red lines show the direction that a ligand atom should be moved to reduce the repulsive energy.
  6. Minimize docking energy
    • With the textport active, type REINITIALIZE (press the return key).
    • Type MINIMIZE_DOCK (press the return key).
    • Select SITE as the molecule to have fixed geometries (press OK).
    • Wait for minimization to finish running.
    • Question 3: What is your final docking energy?
  7. End the docking operation.
    • When finished docking, exit back to the main menu by pressing in the textport.
  8. Warning: If you exit docking mode, then re-enter it, you may get an error message regarding a lack of memory. This is due to a memory allocation problem that cannot be avoided. The only work-around is to exit SYBYL and restart. Freeze the current view and save it to a database if you want to be able to restart exactly where you left off.

 

FlexiDock

Genetic algorithm-based Flexible Docking provides a means of docking ligands into protein active sites. FlexiDock works in torsional space, keeping bond lengths and angles constant. As large vdW interactions can only relax via bond rotation(s), optimization cannot alter chiral centers and bond stereochemistry. FlexiDock works on a protein/ligand pair. The protein backbone atoms are fixed in space, but the ligand is mobile (rotation/translation can be applied). Both the protein (sidechains only) and the ligand can contain a number of flexible bonds. However, to speed up calculations, FlexiDock considers only non-ring single and amide bonds as rotatable.

FlexiDock uses a genetic algorithm to determine the optimum ligand geometry. Genetic algorithms are relatively robust global optimizers, with performance requirements which scale well with increasing system size. The fitness function uses a subset of the Tripos force field: the van der Waals, electrostatic, torsional and constraint energy terms, and calculates the energy of the important atoms in the supermolecule.

  1. Set Up FlexiDock Structures
    • With the ligand docked inside the protein’s active site,
      • Compute >>> FlexiDock >>> Create an Input File.
      • In the Molecule Area dialog, select M1:HIV_protease, then press OK.
        The Set Up FlexiDock Structures, Protein Display Options and No Pocket Defined dialogs are then posted.
      • At the No Pocket Defined dialog, press OK.
    • Protein Display Options dialog.
      • Define and visualize the protein binding pocket.
      • Press Define Pocket
      • Press Substructures and select ASP25, THR26 and GLY27 from both chain A and B and press OK.
      • Increase the radius (Radius to Show Around Pocket) to 10.
    • Set Up FlexiDock Structures dialog.
      • The following seven items in this dialog correspond to the seven steps generally necessary to prepare a protein file from a structural database for submission to FlexiDock, and should be executed in the order suggested. To enforce this the check boxes on the left are organized as a cascade:
        • All are grayed out initially.
        • The first check box becomes accessible only after you have specified the location of the protein and ligand molecules, and both are in separate work areas.
        • Toggling on the first check box activates the next line, and so on.
        • Toggling off any check box turns all the lines below it off and disables them.
      • Five of the check boxes have associated action buttons to Remove the water molecules, Add the hydrogens, Add the atomic charges, Choose the rotatable bonds, and Choose H-bond sites. Each Choose button launches a subsidiary dialog for carrying out the corresponding operation.
        • To perform the desired operation, simply press the button. When the operation is completed, the check box is automatically toggled on and the next line becomes active.
        • If an operation is not necessary, simply toggle on the corresponding check box so you can proceed to the next operation.
      • Two of the check boxes have associated Hint buttons to help you perform the corresponding operations manually: checking the atom types and positioning the ligand in the cavity.
        • The two Hint buttons differ from the four action buttons. Pressing them does not toggle on the associated check box, nor does it activate the next lines.
      • For Water has been removed, press Remove to delete all the water molecules.
      • Check box for Atom types have been checked.
      • For Hydrogens have been added, press Add to fill all valences with hydrogens.
      • For Charges are in place, press Add to compute all atomic charges.
      • For Rotatable bonds are set, press Choose.
        • In the Pick Rotatable Bonds dialog, press All and press OK.
      • For H-bond sites are marked, press Choose.
        • In the Defining H-Bond Sites dialog, press Add, then All and OK for every item.
      • Check box for Ligand is pre-positioned in cavity.
      • Type hiv for file name and press Write Out FlexiDock Input File.
  2. Run FlexiDock.
    • Press FlexiDock It.
      In three successive dialogs you will be prompted to specify:
      • The template parameter file to use. By default, this file is $TA_MOLTABLES/FlexiDock.par.
      • The random number seed to start from.
      • The maximum number for generations to allow. The default is 3000.
    • Accept all default values for the three dialogs.
    • Wait for FlexiDock to finish running.
    • Close Flexidock window by pressing Quit.
    • Delete both protein and drug.
  3. View docking results.
    • Open spreadsheet.
      • File >>> Molecular Spreadsheet >>> New.
      • Select Database and press OK.
      • Select hiv.mdb and press Open.
    • Remove unnecessary information.
      • Select the first row.
      • MSS: Edit >>> Delete.
    • Import list of docking energies.
      • MSS: File >>> Import.
      • Select Custom in the Format option menu of the Import dialog.
      • Press the […] button to its right to bring up the Custom Format dialog.
      • Select Space in the Delimiter option menu and toggle both Labels check boxes on. Enter * and NEW, respectively, in the row and column fields. This means that when the data file is imported its rows will match the existing rows and new columns will be created.
      • Press OK to exit the Custom Format dialog.
      • Enter hiv.mdb/hiv.txt in the File field.
      • Press Import in the Import dialog.
  4. View docked structure.
    • File >>> Database >>> Open.
    • Select hiv.mdb and press Open.
    • Select READONLY as the access mode (press OK).
    • File >>> Database >>> Get Molecule.
    • Select HIV0001 (or select the molecule with the lowest energy).
    • Press OK.
    • Question 4: Is there any difference between the docked conformation and the initial conformation?
    • Question 5: Is there any difference between this docked conformation and the docked conformation from DOCK?
    • Question 6: Is it valid to compare the energies of the docked conformations from DOCK and FlexiDock? Why?
  5. Delete all molecules.
    • Build/Edit >>> Zap (Delete) Molecule.
    • Press All and press OK.

 

Answer to Questions

Question 1: Where is the active site of the protein?
Answer:
protein_active_site.jpg

Question 2: What is the total energy of the minimized structure?
Answer: 36.161 kcals/mol

Question 3: What is your final docking energy?
Answer: A good docked conformation should have negative docking energy.
dock1.jpg

Question 4: Is there any difference between the docked conformation and the initial conformation?
Answer: Before you can answer this question, you must first find a method to detect differences (if any) between the conformations. One method is to measure the distances between several key atoms in the protein and ligand and compare these distances. For example, you might measure the distance between a hydrogen bond donor on the protein and a hydrogen bond acceptor on the ligand and check to see how this distance has changed between the initial conformation and docked conformation.

Question 5: Is there any difference between this docked conformation and the docked conformation from DOCK?
Answer: The approach is the same as the answer given in question 4.

Question 6: Is it valid to compare the energies of the docked conformations from DOCK and FlexiDock? Why?
Answer: No, it is not valid to compare the energies given by the two algorithms. There are two reasons.

  1. The two algorithms may be using different force fields to compute the interaction energy between the protein and ligand.
  2. The atoms in the protein and ligand used to compute the interaction energy is different in the two algorithms. For the DOCK algorithm, only atoms in the box is considered. For FlexiDock, only atoms inside the defined pocket (which is a sphere) is considered.
Share This

This entry was posted by Yap Chun Wei on Saturday, February 16th, 2008 at 11:42 pm and is filed under Pharmacy, Tutorial. You can follow any responses to this entry through the RSS 2.0 feed. You can leave a response, or trackback from your own site.

Leave a Reply

Close
E-mail It