Escolar Documentos
Profissional Documentos
Cultura Documentos
it
http://cassandra.bio.uniroma1.it/
http://www.mmvsl.farm.unipi.it/
1
Introduction
Why autodock?
Because is free under GPL! (http://autodock.scripps.edu/) And because is the most used program
for molecular docking!
As in every science field any experimental methodology needs to be validate prior any production
work!
Therefore in the very first use of the autodock program you will be trained to see if a docking
program (Autodock 4.0.1) could be suitable to study the binding mode of a certain ligands (docking
assessment).
Of course for any docking program the goal should be the reproduction of the experimental bound
conformation of a ligand into its target macromolecule (docking assessment).
Since the HDAC1 (class I HDAC) homology model was prepared starting from the experimental
HDAC8/TSA complex (PDB entry code 1t64) we will use this complex to for the docking
assessment. Unfortunately the HDAC4 template was not co-crystallized with any inhibitor/substrate
therefore the docking validation cannot be performed for class II HDACs
Once we have proven that autodock is able to reproduce the experimental complex the same
procedure can be applied to the modeled HDAC1 and HDAC4 complexes firstly by re-docking the
co-modeled TSA and secondly other ligands (which we know are selective inhibitors) prepared
from a web molecular editor.
NOTE. This tutorial is not to demonstrate the use of autodock4 but just to illustrate the direct us of
the autodock program as potential tools to dock inhibitors in protein models (HDAC1 and HDAC4)
To get more information on the chimera, ADT and autodock program the user should refer to the
respective manuals/tutorials.
2
DOCKING FLOWCHART
(using chimera, ADT and autodock)
3
Get the complex (CPLX) coordinates (i.e. from the PDB)
NOTE. In your home folder you will find the DOCKING folder under
which are saved all the calculations already done for you.
A second folder (EWDD_MD_Tut) has been made in which you find
only the initial files to run all over the tutorial.
4
1. Docking Assessment using the bound ligand conformation
In this tutorial you will be guided in running docking experiments from the AMBER optimized
complex. The program chimera will be used to prepare the macromolecule (lock) and the inhibitor
(key) files. Next the program AutoDockTools 1.4.4 (ADT) will be used to prepare the needed file
and parameters to run the dockings and to analyze the results.
In this first step we will see if the docking program will be successful in reproducing the
experimental complex using as starting point the experimental ligand binding conformation as
found in experimental complex (1t64).
1.1.1. Open chimera by typing chimera at the command prompt of a unix shell. You should always
start chimera in the same directory as the complex file.
5
1.1.2. In the File menu choose Open and select 1T64-A_Min.pdb in the upcoming window. Then
click on the Open button and the structure will appear in the chimera main window.
Do not rotate/traslate the structure in the chimera window during the file preparation!
6
1.1.3. Inspect the 1T64-A_Min.pdb file using a text editor such as vi or whatever you prefer to use.
Check the residue name of the inhibitor (here is named INH) at the end of the file. Using the Select
menu in the Residue sub-menu select the INH residue.
7
1.1.4. Creation of the macromolecule (lock) file. Open the command line option under the
Favorites menu and delete the selected INH residue by typing delete selection in the command line
box. Then save the enzyme (lock) by entering in the command line box:
Check the written file by typing the ls command in the unix shell.
IMPORTANT!!
If the complex comes directly from the AMBER program also the HIE and HID residue have to be
fixed into HIS, otherwise ADT (next section) will not recognise them correctly.
To see how to fix the histidines go to the end of this section (1.1) in the shortcuts.
8
1.1.5. Creation of the ligand (key) file. Close the session and repeat the 1.1.4 point.
Do not rotate/traslate the structure in the chimera window during the file preparation!
9
1.1.6. Select everything but the inhibitor by entering in the command line box the following
command:
And save the inhibitor (key) similarly as seen for the macromolecule.
10
1.1.7. Check the file by loading them all in a new chimera session.
Now the structure can be rotated and translated! (they are already saved)
11
Shortcut version of section 1.1:
The lock and key files can be prepare direcly in a unix shell using some simple unix commands:
Once checked out the inhibitor residue name (see 1.3) the lock and key file can be prepared using
the cat and grep unix commands as following:
or
Prompt> cat cplx_filename.pdb | grep -v INH | sed 's/HIE/HIS/' | sed 's/HID/HIS/' > lock_filename.pdb
12
1.2. preparing the file for docking using ADT and run the docking.
A) You should always start ADT in the same directory as the macromolecule and ligand
files. You can start ADT from the command line in a Terminal by typing "adt" and
pressing <Return> or <Enter>.
B) For both the macromolecule and the ligand, you should always add polar hydrogens,
compute Gasteiger charges and then you must merge the non-polar hydrogens. Polar
hydrogens are hydrogens that are bonded to electronegative atoms like oxygen and
nitrogen. Non-polar hydrogens are hydrogens bonded to carbon atoms.
C) You need one AutoGrid map for every atom type in the ligand plus an electrostatics map.
E.g.: for ethanol, C2H5OH, you would need C, OA and HD maps plus an electrostatics 'e'
map plus a desolvation 'd' map.
D) The grid volume should be large enough to at least allow the ligand to rotate freely, even
when the ligand is in its most fully-extended conformation.
13
1.2.2. Preparing a ligand file for AutoDock.
Start ADT from a unix shell and open a ligand file using the Ligand Æ Input Æ Open … sequence.
Set the file type to *.pdb and choose the key file (1t64_key.pdb). Click OK in the upcoming
window.
14
Save the file as pdbqt (Ligand Æ Output Æ Save as PDBQT…) giving a proper name
(1t64_key.pdbqt) and check the written file.
15
1.2.3.Preparing the macromolecule file.
Open a macromolecule file using the Grid Æ Macromolecule Æ Open … sequence.
Set the file type to *.pdb and choose the lock file (1t64_lock.pdb). Click OK in the upcoming
window. Ignore the warning about the charge and save the file with a proper name
(1t64_lock.pdbqt), ignore the Zn zero charge warning and click OK.
16
In a unix shell edit the 1t64_lock.pdbq file and correct the Zn charge into +2.0.
17
1.2.4. Preparing the GRID parameter file and running Autogrid4.
The grid parameter file tells Autogrid4 which receptor to compute the potentials around, the types
of maps to compute and the location and extent of those maps. and may specify a custom library of
pair-wise potential energy parameters.
18
1.2.4.2. Setting the grid box.
The central position and size of the grid docking box is set using the following procedure:
Set the center of the grid in the center of the ligand (key) and zoom out in the ADT main window
(to zoom out ctrl+middle mouse button or ctrl+c and n keys for the full view) to check the size of
the grid you are making. Turn off the lock molecule display (Display Æ Show/Hide molecule ..)
and re-center the view (ctrl+c and n keys).
As you can see the TSA is almost fully embedded in the grid, but the size is not big enough to allow
a free rotation of the molecule. As a rule of thumb the grid size should be as large as at least twice
the double of the maximum distance you can measure between any two atoms of the co-crystallized
ligand. In the case you do not have any ligand position information the grid should be centered in
the putative binding site and sized to embrace all the residue making the binding pocket.
To set up the grid size you can inspect it visually and adjust the dimension by using the 3
thumbwheel widgets. So for instance the suitable grid size should be 54 x 58 x 74 points using the
default grid spacing of 0.375. Adjust all the values and save the information, save the gpf file and
check the written file.
(Grid option): File Æ Close saving current
19
(Autodock Tools): Grid Æ Output Æ Save GPF…
20
21
Example of a grid parameter file (1t64.gpf)
22
1.2.4.3. Run autogrid4 to make the grid maps. You can easily start a job from the command line
giving the following:
After a few minutes the job will stop and a “done” message will return. Check the list of the file and
you will notice that a number of new files have been created; those are the grid map files that
autodock4 will use for the docking.
And select 1t64_lock.pdbqt and 1t64_key as the macromolecule and ligand filenames, respectively.
Accept all the suggested setting in the AutoDpf4 Ligand Parameters window
23
24
Now let’s set the searching and docking parameters; we will use GALS and for this exercise all the
default value will be let unchanged (only the ga run will be increase to 100). Save the file as
1t64.dpf and check the written files.
25
Docking Æ Output Æ Lamarckian GA…
26
Example of a docking parameter file (1t64.dpf)
27
1.2.5.2. Run autodock4 to launch the docking. You can easily start a job from the command line
giving the following:
After some minutes with the set docking parameters the job is over and a “done” message appears.
Using a modern CPU the job will take about 10-15 minutes to stop (it has taken 11 mins and 39
secs on a AMD Athlon 3200 MHz). So in the meantime go to the Autodock web page and have a
look to the available tutorials in which more detailed information are given
(http://autodock.scripps.edu/).
28
Shortcut version of section 1.2.
All the steps (and files) in section 1.2 can be done (prepared) in a unix shell using in the command
line some stand alone ADT python programs and one awk script utility (pdbbox) in the following
sequence:
Prepare the pdbqt file
Prepare the gpf file for any possible ligand’s atoms and run autogrid4
HINT!
Once all the command have been checked and everything work using a simple unix script is
possible to create a single command that goes directly through sections 1 and 2.
NOTE: the gridcenter in the shortcut is slightly different from that calculated in the ADT section
(1.2.4.2), because ADT calculated the ligand atom closer to the molecular center of mass, while
here the pdbbox utility gave just the molecular center of mass.
29
1.3. Analyzing the docking results.
Reading a docking log or a set of docking logs is the first step in analyzing the results of docking
experiments. (By convention, these results files have the extension ".dlg".)
During its automated docking procedure, AutoDock outputs a detailed record to the file specified
after the -1 parameter. In our example, this log was written to the file 'ind.dlg'. The output includes
many details about the docking which are output as AutoDock parses the input files and reports
what it finds. For example, for each AutoGrid map, it reports opening the map file and how many
data points it read in. When it parses the input ligand file, it reports building various internai data
structures. After the input phase, AutoDock begins the specified number of runs. It reports which
run number it is starting; it may report specifics about each generation. After completing the runs,
AutoDock begins an analysis phase and records details of that process. At the very end, it reports a
summary of the amount of time taken and the words 'Successful Completion'. The level of output
detail is controlied by the parameter "outlev" in the docking parameter file. For dockings using the
GA-LS algorithm, outlev 0 is recommended.
The key results in a docking log are the docked structures found at the end of each run, the energies
of these docked structures and their similarities to each other. The similarity of docked structures is
measured by computing the root-mean-square-deviation, rmsd, between the coordinates of the
atoms. The docking results consist of the PDBQT of the Cartesian coordinates of the atoms in the
docked molecule, along with the state variables that describe this docked conformation and position.
(taken from the ADT online tutorial)
Before starting this section, you should undisplay any molecules in the Viewer using the
30
(click OK if a windows like this appears after clicked on the opne button)
31
1.3.3. Clustering the results.
An Autodock docking experiment usually has several solutions. The reliability of a docking result
depends on the similarity of its final docked conformations. One way to measure the reliability of a
result is to compare the rmsd of the lowest energy conformations and their rmsd to one another, to
group them into families of similar conformations or "clusters".
The dpf keyword, analysis, determines whether clustering is done by AutoDock. As you will see
below, it is also possible to cluster conformations with ADT. By default, AutoDock clusters
docked results at 0.5Å rmsd. This process involves ordering all of the conformations by docked
energy, from lowest to highest. The lowest energy conformation is used as the seed for the first
cluster. Next, the second conformation is compared to the first. If it is within the rmsd tolerance, it
is added to the first cluster. If not, it becomes the first member of a new cluster. This process is
repeated with the rest of the docked results, grouping them into families of similar conformations.
(taken from the ADT online tutorial)
First we will examine the AutoDock clustering that we read in from 1t64.dlg file. Next we will make
new clusterings at different rms tolerance values.
Opens an instance of a Python object, an interactive histogram chart labelled 'ind_out_l:rms = 0.5
clustering'. This chart has bars which represent the clusters computed at the specified rmsd. The
bars are sorted by energy of the lowest-energy conformation in that cluster and start off colored
blue. The height of the bar represents how many conformations are in that cluster. Clicking on a bar
makes that cluster the current sequence for the ligand's CP, and its color changes to red
32
1.3.3.2. Changing the clustering.
Opens a widget which lets you enter a series of new rms tolerances as floating point number
separated by spaces. These will be used to perform new clustering operations on the docked results.
The time consuming step in clustering is computing a difference matrix between conformations to
be compared. Larger rms values require fewer comparisons; conformations which are more similar
require fewer comparisons. If you type a name in the OutputfileName: entry, a clustering output
file will be written. Our convention is to use the extension ".clust" for these files
It is important to set the ligand to the original, input conformation (numbered 0) before clustering.
Type in a list of RMSD tolerances separated by spaces thus 1.0 2.0 3.0 and click on OK. For our
example, this should be very fast. You can visualize the new clusterings by repeating Step 1.3.3.1
33
Best Dock
Best Cluster
The system seem that reach the convergence with ga_num_eval = 2.5M
34
experimental: green; BD_250k: white; BC_250k: cyan; 2.5M: red; 25M: yellow
35
2. Docking Assessment using a ligand conformation other than
experimental.
In this step we will see if the docking program will be successful in reproducing the experimental
complex using a ligand conformation different from that found in experimental complex (1t64).
All the step are similar to those explained in section 1 except that the starting docking conformation
of the ligand will be prepared using a free molecular editor from the web. The user can also use its
preferred molecular editor (SYBYL, Insight, Maestro and others).
ewdd
ewdd07
36
37
After the docking of the TSA_rcmd.mol2 conformation the lowest energy conformation belonged to
the most populated cluster reproduced reasonably well the TSA experimental binding conformation.
38
experimental: blue; 250k: green; BD_2.5M: red; BC_2.5M: cyan; 25M: yellow
The system is going to converge. Comparing with the previous section results seems that the
docking is sensitive to the starting conformation of the ligand. Nevertheless these are quite good
results. You can try to draw the TSA using another molecular editor like MAESTRO from
Schrodinger a dock it into 1t64_lock using autodock.
39
3. Almost full-flexible docking (the new feature in autodock4).
Inspecting the 1t64 complex it is possible to find that Tyr293 hydroxyl group is involved in a
hydrogen bond with the hydroxamic carbonyl oxygen. What if we allow the flexibility of Tyr293 in
the TSA_rcmd docking?
40
41
42
Shortcut version of section 3.1
All the steps (and files) in section 2 can be done (prepared) in a unix shell using in the command
line some stand alone ADT python programs and one awk script utility (pdbbox) in the following
sequence:
Prepare the pdbqt files
3.2. Preparing the parameter files for autogrid (gpf) and autodock (dpf) and running the autogrid
and autodock programs. To do this you can use the ADT command line utilities as following or
follow sections from 1.2.4 through 1.2.5.2 if you prefer the graphical way using ADT.
determine grid center and points using the pdbbox utility for the key and flexible residue together.
43
3.1. Results
The run at ga_num_eval = 2.5M did converged and these are the graphical and clustering results:
44
4. HDAC Selective Inhibitor
Now let’s consider the selectivity issue. In the literature there are only few papers that take it into
account in the assays procedures. Only two of them contains biological data on inhibition of both
HDAC1 and HDAC1. In these papers 4 molecules resulted selective against either HDAC1 (SK
derivatives) or HDAC4 (MAI derivatives). Docking runs, either with the receptor rigid or with the
TYR298 flexible in HDAC1 (HIS349 in HDAC4) have been performed. You can find the results in
the Selectivity folder.
SK-7041 SK-7068
45
MAI_3F MAI_3Cl
46
5. EXERCISE
FOLLOWING THE STEPS DESCRIBED IN THE TUTORIAL, YOU CAN RE-DOCK THE
TSA IN HDAC_1 AND HDAC_4 MODELS AND STUDY THE BINDING MODE OF THE
SELECTIVE INHIBITORS (SK-7041, SK-7068, MAI_3F and MAI_3Cl) ABOVE
REPORTED.
For question of time you can use the dockings saved in the DOCKING folder and directly analyze
them.
As part of the exercise you are also encouraged to dock different molecules to see if would be
possible to design selective HDAC inhibitors.
These are the calculations we have run that you can find in the DOCKING folder
N key lock receptor Geometry 250k 2.5M 25M
1 1t64 1t64 rigid Min 11m 49s 1h 43m 17h 3m
2 TSA_rcmd 1t64 rigid Min 9m 45s 1h 45m 15h 7m
3 TSA_rcmd 1t64 flex Min 15m 42s 2h 35m X
4 TSA HDAC1 rigid MD 10m 31s 1h 35m 15h 50m
5 TSA HDAC1 flex MD 14m 46s 2h 30m X
6 TSA HDAC1 rigid Min 10m 16s 1h 41m X
7 TSA HDAC1 flex MIn 16m 20s 2h 41m X
8 TSA HDAC4 rigid MD 10m 3s 1h 39m 17h 4m
9 TSA HDAC4 flex MD 13m 25s 2h 25m X
10 TSA HDAC4 rigid Min 10m 13s 1h 41m X
11 TSA HDAC4 flex Min 14m 44s 2h 25m X
12 MAI_3F HDAC1 rigid MD 10m 21 1h 41m X
13 MAI_3CL HDAC1 rigid MD 9m 49s 1h 39m X
14 SK_7041 HDAC1 rigid MD 12m 37 2h 5m X
15 SK_7068 HDAC1 rigid MD 13m 58s 2h 18 X
16 MAI_3F HDAC1 flex MD 18m 4s 2h 59 X
17 MAI_3CL HDAC1 flex MD 15m 6s 3h 3m X
18 SK_7041 HDAC1 flex MD 22m 1s 3h 33m X
19 SK_7068 HDAC1 flex MD 24m 6s 3h 40m X
20 MAI_3F HDAC1 rigid Min 9m 33s 1h 34m X
21 MAI_3CL HDAC1 rigid Min 9m 37s 1h 34m X
22 SK_7041 HDAC1 rigid Min 11m 47s 1h 56m X
23 SK_7068 HDAC1 rigid Min 13m 2h 09m X
24 MAI_3F HDAC1 flex Min 16m 20s 2h 41m X
25 MAI_3CL HDAC1 flex Min 16m 23s 2h 40m X
26 SK_7041 HDAC1 flex Min 19m 16 3h 10m X
27 SK_7068 HDAC1 flex Min 20m 55 3h 27m X
28 MAI_3F HDAC4 rigid MD 9m 20s 1h 32m X
29 MAI_3CL HDAC4 rigid MD 9m 26s 1h 32m X
30 SK_7041 HDAC4 rigid MD 11m 25s 1h 53m X
31 SK_7068 HDAC4 rigid MD 12m 37s 2h 5m X
32 MAI_3F HDAC4 flex MD 15m 13s 2h 31m X
33 MAI_3CL HDAC4 flex MD 13m 44s 2h 32m X
34 SK_7041 HDAC4 flex MD 18m 42s 3h 3m X
35 SK_7068 HDAC4 flex MD 22m 9s 2h 54m X
36 MAI_3F HDAC4 rigid Min 10m 22s 1h 53m X
37 MAI_3CL HDAC4 rigid Min 10m 23s 1h 40m X
38 SK_7041 HDAC4 rigid Min 15m 02s 2h 25m X
39 SK_7068 HDAC4 rigid Min 16m 43s 2h 23m X
40 MAI_3F HDAC4 flex Min 13m 27s 2h 12m X
41 MAI_3CL HDAC4 flex Min 14m 03s 2h 28m X
42 SK_7041 HDAC4 flex Min 16m 12s 2h 40m X
43 SK_7068 HDAC4 flex Min 17m 43s 2h 55m X
N: number of docking; key: ligand being docked; lock: macromolecular target; receptor: how the
receptorwas treated (flex or rigid); geometry: which conformation of the lock was used in the
docking; 250k/2.5M/25M: Docking time at different ga_num_evals.
47
The first three docking are part of the docking assessment in the sections from 1 to 3, while the
remaining docking experiments are the effective studies in the modeled HDAC1 and HDAC4. As
you can see we have performed several dockings using different energy states of the HDAC1 and
HDAC4 complexes (about 90 jobs) . Below are reported the graphical and clustering results (rmsd
= 2) of the docking at ga_num_evals = 2.5M for the docking 4-11 of the above table.
48