Profa.

Rafaela Salgado Ferreira

Departamento de Bioquímica e Imunologia - UFMG

Instruções para Tutorial DockThor

O docking (de um ligante da protease do HIV) será realizada através do webserver do

programa DockThor:
[Link]
Será realizado o procedimento de docking nativo, no qual emprega-se um ligante que
já co-cristalizado com a proteína de interesse e considera-se a conformação proteica
observada na presença deste ligante (obtida diretamente do PDB referente ao complexo em
questão).

Preparação inicial de arquivos:

Utilizaremos o tutorial preparado pela própria equipe do DockThor. No entanto, antes de

iniciar o tutorial é necessário preparar os arquivos da proteína e do ligante a ser avaliado. É
necessário também definir o ponto central do Grid no qual será realizado o procedimento de
docking.

1. Para preparar o arquivo da proteína:

- Faça o download do arquivo [Link] diretamente no site do PDB:

[Link]

Estrutura 1HBV - A CHECK ON RATIONAL DRUG DESIGN. CRYSTAL STRUCTURE OF A COMPLEX

OF HIV-1 PROTEASE WITH A NOVEL GAMMA-TURN MIMETIC

Não é necessária nenhuma preparação adicional. O receptor será preparado

automaticamente no site do programa.

2. Para preparar arquivo do ligante:

Num editor de texto, abra o arquivo [Link] e copie todas as linhas contendo
informações sobre o ligante. Estas são identificadas pela coluna HETATM e pelo código do
ligante GAN.

HETATM 1523 N1 GAN A 600 -4.428 16.771 18.898 1.00 49.56 N

HETATM 1524 C2 GAN A 600 -4.368 17.218 20.328 1.00 48.46 C
HETATM 1525 C3 GAN A 600 -3.362 18.372 20.475 1.00 51.15 C
…
HETATM 1564 C42 GAN A 600 -12.454 13.101 33.796 1.00 39.47 C

- Salve estas linhas em um novo arquivo: [Link].

 3. Definição do ponto central do GRID:

Para definição do ponto central do GRID, selecione um átomo central no sítio ativo. Para
visualizar o sítio ativo, utilize o programa Pymol e crie um objeto contendo o ligante GAN e
todos os resíduos a uma distância de até 6 Å ao redor.

Considerando a posição central dos Asp catalíticos no sítio ativo, neste caso selecionei o
oxigênio OD1 do Asp25 para definição do ponto central do grid. A partir da linha
correspondente a este átomo no arquivo PDB, obtem-se as coordenadas para o GRID
(destacadas em negrito abaixo).

ATOM 198 OD1 ASP A 25 -11.544 21.448 28.812 1.00 10.03 O

Estas coordenadas serão adicionadas no site do DockThor, no momento de submissão

do docking.

Após a preparação destes arquivos, siga o tutorial a seguir, que foi retirado diretamente do site do
DockThor.

Observações importantes:

- Na preparação da proteína, altere o estado de protonação de Asp25 para AspN1

(conforme indicado no próprio tutorial). O pka das cadeias laterais dos resíduos na
estrutura 1HBV foram calculados utilizando o programa propka3.0. Conforme indicado
abaixo, observa-se uma variação significativa no pKa de Asp25, quando comparado a
Asp29:

RESI DUE pKa BURI ED

ASP 25 A 5. 91* 100 %

ASP 29 A 3. 77 51 %

Essa variação é decorrente da proximidade espacial entre dois resíduos de aspartato.

- Nenhum cofator será utilizado.

 Portal

[Link]

User Guide
Version 1.0

Contributors

Diogo A. Marinho, Isabella A. Guedes, Eduardo Krempser, Camila S. de

Magalhães, Helio J. C. Barbosa and Laurent E. Dardenne

[Link]

Laboratório Nacional de Computação Científica – LNCC/MCTI

14 July 2013

INPI Software Registration Number 13318-3

DocktThor® registered mark
 1

Portal DockThor

1. Introduction

The DockThor Portal, developed by the group GMMSB/LNCC, is a free receptor-

ligand docking server idealized to facilitate and enable the use of the docking
methodology by the academic community. The implemented DockThor® program is a
flexible-ligand and rigid-receptor grid based method that employs a multiple solution
genetic algorithm along the MMFF94S molecular force field scoring function. The main
steps of the ligand and protein set up are available on the DockThor Portal, being
possible to change the amino acid residues protonation states and include cofactors
(e.g. structural water molecules, metals, organic molecules) as rigid entities. The user
can also customize the main parameters of the energy grid and the genetic algorithm.
The results of the docking process can be analyzed and sorted automatically.
The analysis parameters can also be customized by the user. The DockThor Portal
employs the computational facilities provided by the Brazilian SINAPAD (Sistema
Nacional de Alto Desempenho) high performance platform.

2. Submiting a Docking Job

In the present version 1.0 of the DockThor Portal only protein receptors and
pdb type files can be accepted (or pre-prepared DockThor input files *.in). Other types
of receptors (e.g. DNA, RNA, another ligand) will be allowed in the next version.

2.1 Protein Preparation

1. Click on the Docking tab. Then click on the Protein tab to open the protein
preparation page.
 2

2. To submit your protein file click on the Upload button. So far, it is only
possible to upload Protein Data Bank (.pdb) type or DockThor input (.in)
protein files.

3. Prepare the protein file applying the basic options clicking on Prepare. At
this step, all the missing atoms of the residue side chains will be
reconstructed. The protein atoms are recognized by the initial .pdb label
'ATOM', Atoms associated with the initial .pdb label 'HETATM' will be
ignored. All the atoms are also recognized by their .pdb atom label (e.g. CA,
CB etc). If the atom label nomenclature is not right the atom will be
reconstructed. For this reason it is very likely that all the hydrogen atom will
be rebuilt1. If two side chain conformations are given for the same residue
in the .pdb file, only the first one will be considered by the program.

1
To maintain the original H's positions you must use the same H atom .pdb nomenclature used
by the DockThor program (please download and examine the generated *.in file).
 3

4. Set the protonation states of the residues (Asp, Cys, Glu and His are set to
the default values - see appendix A). Reprepare your protein file by applying
the new protonation states clicking on Reprepare.

5. View the prepared protein file with JSmol clicking on View 3D.
 4

6. To download the respective protein files click on Download prepared files.

[Link] – DockThor receptor input file (contains the MMFF94S atomic type
number, the atomic coordinates, the bond connectivities, the atomic partial
charges, the pdb type atom, residue and chain labels and the residue and chain
numbers).

[Link] – contains the information of the protein preparation process.

protein_prep.pdb – prepared protein file according with the Protein Data Bank
format (equivalent to the [Link] file for visualization).

protein.X - configuration file for each protein chain X. Contains the amino acid
residue list and the respective protonation state label.

7. Click Next to submit your prepared protein file to docking. It is noteworthy

that only the protein file in the .in format is necessary to docking.
 5

2.2 Ligand Preparation

1. To submit a single small molecule2 to docking, click Upload. The formats

accepted are those recognized by OpenBabel. For a complete list of these
files, click on the (Help) link.

2. The uploaded ligand needs to be prepared to generate the respective

topology file (.top) to docking. This step comprises the right MMFF94S force
field atom type assignment, the atomic partial charges calculation and the
assignment of the rotatable chemical bonds. If you want to add hydrogen
atoms automatically, just check the Add hydrogens box and Reprepare the
molecule again. The hydrogen atoms will be added through OpenBabel tool
at pH = 7.0.

3. To view the current structure of your ligand, click View 3D.

2
In the present portal version, it is only possible to dock one ligand at time. A virtual screening version of
the DockThor Portal will be available in the future.
 6

4. You can view the selected rotatable bonds just clicking on the Rotatable
bonds... button. If you unselect some flexible rotatable bonds click on
Reprepare.
 7

5. If you click on Reprepare without unselecting any rotatable flexible bond,

the original set of flexible bonds will be restored.

6. To download the prepared ligand files click on Download prepared files.

[Link] – DockThor ligand input file. It contains the atom name, the atom
number, the MMFF94S atomic type number, the atomic partial charge, the atomic
coordinates and the atom valence. It contains also: (i) the atom connectivity; (ii)
force field torsional parameters; (iii) selected flexible bonds; (iv) non-bonded
intramolecular atom interactions.

new_ligand.pdb – prepared ligand in the Protein Data Bank format (for

visualization). This file is only generated when the Add hydrogens option is
applied.

7. Click Next to send the updated ligand file to docking.

2.3 Cofactor and Water Preparation

1. For some protein-ligand complexes it is important to consider cofactor (e.g.

NAD, ATP, FAD, Mg, Zn etc.) and/or water/solvent molecules. The DockThor
Portal allows the inclusion of one or more cofactor and solvent molecules;
they are kept fixed during the docking simulation. Upload each file at a time
clicking on Upload.
 8

2. As well as ligand file, the cofactor and water files (one or more water
molecules per file) need to be converted to the topology file. To add
hydrogen atoms3, check the respective boxes and click Prepare.

3. It is also possible to download the prepared files corresponding to each

cofactor/water clicking on Download prepared files. Each cofactor/water
file will generate one link for download.

4. Send the cofactor and/or water to docking by clicking on Next.

2.4 Docking Configuration

1. The next tab displays the main parameters to run the docking job. It is
possible to see the original files for the ligand and protein in Files uploaded.
Check if they are all right.

2. Fill the Email field to receive the link with the results page when the docking
job is finished. This field is mandatory.

3
Since MMFFLigand does not optimize the hydrogen atoms position, it is recommended to do
this previously and do not chose to include hydrogen atoms to these water files.
 9

3. Fill the coordinates of the energy grid center in Grid center (Xc, Yc, Zc).

4. Fill the Grid Dimensions (±Δx, ±Δy, ±Δz). These values correspond to half of
the grid size on each dimension, e.g. (Xc-Δx) ≤ X Dimension ≤ (Xc+Δx)

5. Select the spatial discretization of the energy grid. This value corresponds to
the spacing between the grid points (the default value is 0.25Å). Check if the
number of grid points does not exceed the limit allowed (1,000,000) in the
Grid points box.

6. To view the grid dimensions and the receptor click View Grid in 3D.

7. Choose an identification label for your docking job.

8. Some genetic algorithm (GA) parameters can be modified.

 10

- The GA multisolution algorithm was optimized to deal with highly flexible

ligands. We do not recommend changing the standard number of evaluations
per run (i.e. 1,000,000) or the GA population size (i.e. 1,000). Change these
values at your own risk.

- Each submission job corresponds to 30 independent docking runs. A maximum

of 50 docking runs per job is permitted. It is possible to run more independent
docking runs submitting more than one job and changing the initial seed. The
seed must have a negative value, and for each successive run its value is
diminished by one. You should take this into account if you want to submit
more than one job in order to obtain more than 50 independent docking runs.

9. Run your docking job by clicking on Dock!

 11

10. When your docking process is finished an email will be sent at the
previously stated e-mail, containing a link to the results page.

3. Analyzing the Docking Job

Once the docking job is finished, the user will receive an email with the link to
the Results page as below:
Your docking result is available at
[Link]

Follow the link in the email to access the Results and Analyses page.

3.1 Results and Analyses

For each docking run the final population is clustered using the total energy and
a RMSD = 1.0Å criterion. Only the leaders of each cluster will be used in the final
clustering analysis step:

1. It is possible to cluster and sort out the docking solutions according to two
criteria: Total Energy (intermolecular ligand-receptor + intramolecular ligand
energies) or Interaction Energy (only intermolecular ligand-receptor
energy). The default is Analysis by total energy.
 12

2. Click Analyze to perform the docking poses analyses.

3. You can change the clustering criterion (default = 2.0Å) and the number of
the best ligand binding modes (default = 20) selected for analyses and
visualization. Click on the Result Settings button.
 13

4. It is possible to view the results interactively on the website. Click on View

results interactively.

5. To see each docking ligand binding mode (i.e. cluster leader) select the
corresponding one in the results table. For each solution the system shows
the corresponding Run of the Genetic Algorithm, the Model (number of the
cluster leader) of the corresponding GA run, the Total Energy, the
Intermolecular Energy and the RMSD (root mean square deviation
calculated using the non H atoms) relative to the top ranked pose. This value
gives an idea of the conformational difference among the alternative ligand
binding modes.
 14

6. It is also possible to upload a reference ligand conformation (.pdb type file)

to perform redocking analyses or help the investigation of the distinct
binding modes. Then click Analyze to perform the docking poses analyses
and click on View results interactively. The RMSD (in the last column) is
now calculated relative to the uploaded reference ligand pose. It is possible
to hide the protein structure to facilitate visualization.
 15

7. To download the docking results, including the clustered conformations,

click Download Results.

[Link] – general information about each DockThor run.

[Link] – general information about the docking parameters used.

[Link] – total number of clusters obtained (after analyzing all runs).

[Link] – summary of the clustering analysis for the best cluster leaders (ranking,
energies, RMSD).

out.mol2 – contains the atomic coordinates of the best cluster leaders (multimodel
.mol2 type file, sorted out according to the results described in [Link]).

ligand_run_X.log – contains the information of the cluster leaders obtained in run X

(using a RMSD criterion of 1.0Å).

ligand_run_X.pdb – contains the atomic coordinates of the cluster leaders obtained in

run X (multimodel .pdb type file).
 16

[Link] – DockThor receptor input file.

[Link] – DockThor ligand input file.

8. It is always possible to perform other analyses and download the respective

files.

4. Softwares

The Portal DockThor uses the following programs:

MMFFLigand: generates the topology file for the ligand and cofactor files
through MMFF94S force field and OpenBabel tools;

PdbThorBox: prepares the protein file (adds hydrogen atoms, changes amino
acid protonation states, completes missing side chains) with the MMFF94S
force field;

DockThor: the docking program is a flexible ligand rigid receptor grid based
method that employs a multiple solutions genetic algorithm as the search
method using the MMFF94S force field as the scoring function.

Dtstatistic: clusters and ranks the docking poses according to total or

interaction energies.

* All these programs were developed by the GMMSB/LNCC group.

** Ernesto R. Caffarena, Michel Loos and Isabelle Ortmans also contributed for the
PDBTHORBOX development.
 17

Appendix A

Residue Protonation States

ASP - Negatively charged aspartic acid (default).

ASPN1 - Neutral aspartic acid with a H bonded to the Oδ1
ASPN2 - Neutral aspartic acid with a H bonded to the Oδ2

GLU - Negatively charged glutamic acid (default).

GLUN1 - Neutral glutamic acid with a H bonded to the Oε1.
GLUN2 - Neutral glutamic acid with a H bonded to the Oε2.

CYSH - Neutral cysteine with a H bonded to S (default).

CYS - Negatively charged cysteine.
CYSS - Neutral cysteine (disulfide bond).

HIS - Neutral histidine with a H bonded to N (default)

HISD - Neutral histidine with a H bonded to N.
HISP - Positively charged histidine.

ARG - Positively charged arginine (default).

ARGN1 - Neutral arginine at Nω1.
ARGN2 - Neutral arginine at Nω2.

LYS - Positively charged lysine (default).

LYSN - Neutral lysine.
Análise dos resultados:

No site do DockThor é possível visualizar os resultados, através de link enviado por

e-mail. Resultados de uma submissão prévia podem ser acessados pelo link:

[Link]
16_ligand_iwgvow

É possível comparar os resultados com o ligante de referência (desde que você

acrescente o arquivo contendo a conformação cristalográfica antes de clicar em Analyze).
Observe se o modo de ligação de menor energia (colorido por átomos) corresponde ao modo
cristalográfico (em verde).

Para analisar cuidadosamente os resultados, abra no Pymol os arquivos:

- protein_******_prep.pdb (arquivo da proteína após preparação pelo DockThor)

- [Link] (ligante com coordenadas cristalográficas)
- out.mol2 (conformações de menor energia)

Altere as representações gráficas das proteínas e dos ligantes da maneira que achar
mais conveniente para análise. Sugestão de etapas para a análise:

- crie um objeto contendo os resíduos presentes a uma distância de no máximo 6Å do

ligante cristalográfico. Represente este objeto como sticks, colorindo por átomo.
- represente a proteína como cartoon
- represente o ligante [Link] e os resultados de docking como sticks, colorindo
ambos por átomo, porém colorindo de cores diferente os carbonos do ligante cristalográfico.
- no menu Edit > Edit all , altere o valor de stick_radius de 0.25 para 0.15 e dê enter.
 Após todas estas etapas você deverá visualizar uma tela semelhante a mostrada
abaixo:

A seguir, alterne entre as conformações de menor energia calculadas por docking,

utilizando o botão indicado na figura abaixo:

Meça as distâncias das ligações de hidrogênio entre o ligante cristalográfico e a

proteína. Observe se estas interações são preditas nos modos de ligação calculados por
docking.