Template for OpenGL C scripts.

This script was generated automatically by Tree2C and performs the classification of molecules between permeant and not permeant of blood-brain barrier (BBB) through a decision tree. For more information, click here.

MetaClass predictor.vll

MetaClass is a comprehensive classification system for the prediction of the metabolic reactions of a given molecule or of a set of molecules. The prediction is based on a machine-learning algorithm (Random Forest), which was trained by using the metabolic data collected and classified into the MetaQSAR database. MetaClass package includes two modules: MetaClass builder and MetaClass predictor. The former can generate automatically not only the models but also code, which can be used directly in the VEGA ZZ environment for the prediction, while the latter is the result of MetaClass builder run.

Software requirements for MetaClass predictor:

• MOPAC 2016 (freely available after registration at http://openmopac.net). For the version selection and the right installation, you must consult click here.

Software requirements for MetaClass builder:

• The same packages required for MetaClass predictor.

• Weka 3.8 or greater (freely available at https://www.cs.waikato.ac.nz/ml/weka/index.html).

• Java (freely available at https://www.java.com, required to run Weka).

• Tree2C (included in VEGA ZZ package).

• Tcc C compiler (included in VEGA ZZ package).

• A MetaQSAR database in the supported format (ODBC data source, SQLite and Microsoft Access).

MetaClass predictor

The MetaClass predictor is a compiled C-Script (available for both x64 and x86 architectures), which allows the prediction of metabolic reactions which a given molecule undergoes according to MetaQSAR rules. To run it, you must start VEGA ZZ, select File → Run script in the main menu, expand the ADMET branch and double click MetaClass predictor.vll. If a molecule is present in the current workspace, the prediction is performed for a single molecule and the results are shown in the VEGA ZZ console. If the workspace is empty, a file requester is shown to select an input database (it must be in a format supported by VEGA ZZ: Microsoft Access, Mol2, ODBC data source, SDF, SMILES, SQLite and Zip). In this second case, the prediction is performed for all molecules in the database and the results are saved into a CSV file. Since the training set used by the learning phase (the substrates classified in the MetaQSAR database) includes only molecules in non-ionized form, with the exception of quaternary ammonium salts, the molecules for which you want to predict the metabolic reactions must also be in their neutral form. Here is the typical output shown by the VEGA ZZ console for a single molecule prediction (eg. Naproxen):

* Prediction of MetaQSAR metabolic reactions
* Assigning atom charges
  Total charge: 0.00

   Reaction Substrate Dom. viol.
  ======================================================================
   01 - Oxidation of Csp3 Yes 0
   02 - Oxidation of Csp2 & Csp Yes 0
   03 - -CHOH <-> >C=O -> -COOH Yes 0
   04 - Various redox reactions of carbon atoms No 0
   05 - Redox reactions of R3N No 0
   06 - Oxidation of >NH, >NOH and -N=O // Reduc Yes 0
   07 - Oxidation to quinones or analogs // Redu Yes 0
   08 - Oxidation and reduction of S atoms No 0
   09 - Redox reactions of other atoms Yes 0
   11 - Hydrolysis of esters, lactones and inorg Yes 0
   12 - Hydrolysis of amides, lactams and peptid No 0
   13 - Epoxide hydration No 0
   14 - Other hydrolysis/hydration reactions // No 0
   21 - O-Glucuronidations & glycosylations No 0
   22 - N- and S-Glucuronidations // All other g No 0
   23 - Sulfonations (O-, N-, ...) Yes 0
   24 - GSH & RSH conjugations + sequels // GSH- No 0
   25 - Acetylations & acylations No 0
   26 - CoASH-Ligation followed by amino acid co Yes 0
   27 - Methylations (O-, N-, S-) No 0
   28 - Other conjugations (PO4, CO2, ...) // Tr No 0

For each reaction class (according to the metabolic reaction classification in MetaQSAR), the output table shows the code and the description (Reaction column), if the molecule is substrate or not (Substrate column) and the number of the domain violations (Dom. viol. column). This counter indicates how many parameters/attributes are out of the range of the property space of the training set. If this value is not zero, the prediction might be less accurate.
If the input is a database of molecule, the output is saved to a CSV file thanks to a file requester, which allows you to choose the file name. The prediction can be aborted in any time just clicking the Abort button shown in the progress window. The output file includes a column with the name of the molecule for which the prediction is given, and two columns for each reaction class reporting respectively the prediction (1 if the molecule is substrate, 0 if it is not) and the number of domain violations.

MetaClass builder
As explained above, MetaClass builder can generate the predictive models based on MetaQSAR dataset as well as the C-Script code required to build the MetaClass predictor. To run it, you must start VEGA ZZ, select File → Run script in the main menu, expand the ADMET branch and double click MetaClass builder.c. The only input required by the script is a database, which must be in a format supported by VEGA ZZ (ODBC data source, SQLite and Microsoft Access).
MetaClass builder modifies the database structure by adding three new tables, which include the descriptors/attributes calculated with MOPAC and Kier-Hall approach (namely Prop_Mopac, Prop_Mopac_Atom and Prop_KierHall). Therefore, if you want to keep the original structure, you must create a work-copy of the database. Moreover, if the script finds these tables, the calculation of the descriptors is not performed so speeding-up the whole process. If you want to force the descriptors calculation, you should delete these three tables from the database.

How MetaClass builder works
MetaClass builder uses a multi-step approach to build the final models, which are compiled as link-libraries that can be used directly in VEGA ZZ environment as shown here:

1. Checking the required programs/components. If one of them is missing, the script aborts.

2. Checking if the input file/data source (specified by the file requester) is a MetaQSAR-compatible database.

3. Extracting the reaction classes for which the predictive models will be developed.

4. Calculating and storing into the same database the Kier-Hall descriptors if required.

5. Calculating and storing the MOPAC descriptors if required by applying the following keywords: PM7 GEO-OK MMOK 1SCF SUPER THREADS=1. If your molecules are not optimized by MOPAC, you must delete the 1SCF keyword from VGS_MOPAC_KEYS definition of the script.

6. Extracting the VEGA-based molecular descriptors from the database, which are calculated by MetaQSAR when the user compile it.
   
   For each reaction class:

7. Building for each class a balanced dataset by selecting all substrates of the reaction class and an equal number of non-substrates (which are chosen randomly).

8. Creating the input file for Weka (in ARFF format) by considering the most significant attributes previously chosen by Select attributes tool implemented in Weka (see ADMET\_MetaClass builder\*.txt files) according to the BestFirst search algorithm (direction = Forward; lookupCacheSize = 1; searchTermination = 5) and the WrapperSubsetEval attribute evaluator (classifier = RandomForest with default settings; doNotCheckCapabilities = False; evaluationMeasure = accuracy, RMSE; folds = 5; seed = 1; threshold = 0.01).

9. Running Weka to build the models by Random Forest algorithm with the default parameters, which are: weka.classifiers.trees.RandomForest -P 100 -print -I 100 -num-slots 1 -K 0 -M 1.0 -V 0.001 -S 1.

10. Checking if the Weka calculation is completed without errors and reads the output file to extract the statistical data.

11. Running Tree2C to convert the decision trees generated by Weka into C-Script code.

12. Compiling the C code (for both x64 and x86 versions) into the object file by using Tcc.
    
    For all reaction classes:

13. Creating the configuration header file of the main code according the data collected during the generation of the models.

14. Copying the main code template (main.c) from ADMET\_MetaClass builder directory to the working directory, compiling and linking it by Tcc with the object files created for each reaction class.

15. Installing the resulting compiled scripts (.vll and .vl1) into the ADMET scripts directory of VEGA ZZ program.

16. Cleaning the working directory if VGS_CLEANUP macro is defined in the code of the script (by default this is not performed).

The MetaClass builder generates both x64 and x86 versions of the MetaClass predictor only if both VEGA ZZ x86 and x64 are installed. Usually, only one of the two versions is installed according to your operating system, but you can override this behaviour during the VEGA ZZ setup by choosing the installation of the Live CD creator component.
The working directory is the same in which the MetaQSAR database file is placed and here several intermediate files are saved. In detail, you can find:

• config.h: the configuration header of the main part of MetaClass predictor (main.c).

• DATABASE_NAME - Model performances.csv: this file includes several statistical data of the models created by Weka (see below).

• main.c: the main code of MetaClass predictor.

• main_32.o: the object file compiled by Tcc from main.c (x86 version).

• main_64.o: the object file compiled by Tcc from main.c (x64 version).

• MetaClass predictor.vll: the final script compiled and linked by Tcc (x86 version).

• MetaClass predictor.vl1: the final script compiled and linked by Tcc (x64 version).

For each reaction class you can find:

• DATABASE_NAME – REACTION_CODE.arff: the Weka input file in ARFF format.

• DATABASE_NAME – REACTION_CODE.txt: the Weka output file with the trees in text format.

• model_REACTION_CODE.c: the source code of the decision tree translated by Tree2C.

• model_REACTION_CODE.o: the object file of the decision tree compiled by Tcc (x86 version).

• model_REACTION_CODE_32.o: the object file of the decision tree compiled by Tcc (x86 version).

• model_REACTION_CODE_64.o: the object file of the decision tree compiled by Tcc (x64 version).

As explained above, DATABASE_NAME - Model performances.csv includes several statistical data about the performances of the models built by Weka:
 

This script was generated automatically by Tree2C and performs the classification of molecules between mutagen and not mutagen through a decision tree. For more information, click here.

It calculates the dipole momentum by AMMP. If the charges aren't assigned, they are fixed by Gasteiger - Marsili method (see AMMP's DIPOLE command).

Vnonbon internal lys.n 137 Eq -12.860423 E6 -1.397398 E12 2.191076
Vnonbon external lys.n 137 Eq 16.632879 E6 -5.806829 E12 9.177713
Vnonbon total lys.n 137 Eq 3.772456 E6 -7.204227 E12 11.368790

where internal is intramolecular energy, external is the intermolecular (interaction) energy, total is the sum of intramolecular and intermolecular energies, Eq is the electrostatic (coulombic) energy, E6 and E12 are the Lennard - Johnes terms. At the end of the atom dump, AMMP shows also:

Vnonbon total internal 151.439880
Vnonbon total external 2.272158
Vnonbon total 153.712067
153.712067 non-bonded energy
153.712067 total potential energy

where Vnonbon total internal is the total intramolecular energy, Vnonbon total external is the total intermolecular (interaction) energy, Vnonbon total is the total non-bond interaction energy (it's the sum of Vnonbon total internal and Vnonbon total external). Non-bonded energy and total potential energy are self explaining.

Finally, the results (Vnonbond total internal, Vnonbond external and Vnonbond total) are copied to the clipboard.

The AMMP's Kohonen neural network is used to find the 3D space filling curve corresponding to the structure.  If the charges aren't assigned, they are fixed by Gasteiger - Marsili method (see AMMP's KOHONEN command).

It performs a rigid rocking calculation by genetic algorithm as implemented in AMMP program. This calculation requires two molecules in the workspace: the first one must be the receptor and the second one must be the ligand. This last molecule is moved to obtain the complex. Both molecules must have the hydrogens and the charges are automatically fixed (Gasteiger - Marsili method) if they are unassigned.
This script has a graphic user interface (provided by GraphApp library) and to understand the meaning of each field, it's strongly recommended to read the GDOCK documentation.

REMARK 300
REMARK 300 BIOMOLECULE: 1
REMARK 300 THIS ENTRY CONTAINS THE CRYSTALLOGRAPHIC ASYMMETRIC UNIT
REMARK 300 WHICH CONSISTS OF 2 CHAIN(S). SEE REMARK 350 FOR
REMARK 300 INFORMATION ON GENERATING THE BIOLOGICAL MOLECULE(S).
REMARK 350
REMARK 350 GENERATING THE BIOMOLECULE
REMARK 350 COORDINATES FOR A COMPLETE MULTIMER REPRESENTING THE KNOWN
REMARK 350 BIOLOGICALLY SIGNIFICANT OLIGOMERIZATION STATE OF THE
REMARK 350 MOLECULE CAN BE GENERATED BY APPLYING BIOMT TRANSFORMATIONS
REMARK 350 GIVEN BELOW.  BOTH NON-CRYSTALLOGRAPHIC AND
REMARK 350 CRYSTALLOGRAPHIC OPERATIONS ARE GIVEN.
REMARK 350
REMARK 350 BIOMOLECULE: 1
REMARK 350 APPLY THE FOLLOWING TO CHAINS: B, A
REMARK 350   BIOMT1   1  1.000000  0.000000  0.000000        0.00000
REMARK 350   BIOMT2   1  0.000000  1.000000  0.000000        0.00000
REMARK 350   BIOMT3   1  0.000000  0.000000  1.000000        0.00000
REMARK 350   BIOMT1   2 -1.000000  0.000000  0.000000      174.00000
REMARK 350   BIOMT2   2  0.000000 -1.000000  0.000000      174.00000            
REMARK 350   BIOMT3   2  0.000000  0.000000  1.000000        0.00000

To build this homodimeric macromolecule:

• Open the original PDB file.
• Run Coordinate transformation script,
• Put in the dialog window the values shown in red.
• Click Apply button.
• Reopen the original PDB file in the same workspace of the transformed structure and click Append in the dialog window.

This scripts build a single-walled carbon nanotube (SWCNT) structures. It's based on VBS code developed by Roberto G. A. Veiga at Instituto de Física - Universidade Federal de Uberlândia (UFU) - Brazil, using the algorithm described in the article: White et al., Phys. Rev. B, 1993, Vol. 47, No. 9, pp. 5485-88.

ResName:ResNum:ChainID:MolNum List_Of_Aminoacids

where ResName is the name of the residue (max. 4 characters), ResNum is the residue number (max. 4 characters), ChainID is the chain identifier (1 character), MolNum is the molecule number (non zero, unsigned integer) and List_Of_Aminoacids is the list of the aminoacids that will be sequentially replaced (max. 20 characters, aminocid single character code). ChainID and MolNum are optional parameters, but it you want to specify the molecule number without to indicate the chain, you can use * as ID. # and ; at the beginning of each line can be used for remarks.

Example:

; Mutation list example

THR:3 EYF
SER:6:Y AL

It generates 6 mutants, involving the residues in 3 and 6: EA, YA, FA, EL, YL and FL.

Each mutated protein is automatically minimized by NAMD 2 (5000 steps), keeping the backbone fixed.

WARNING:
To run this script, NAMD 2 package and parm.prm parameter file must be installed. For more detail, click here and here.

It moves the atoms at the specified coordinates. Checking Active atoms only, only the visible atoms are moved.

Lipinski's rule of five
This rule establishes that  an orally active drug must have:

• not more than 5 hydrogen bond donors (nitrogen or oxygen atoms with one or more hydrogen atoms);
• not more than 10 (2 x 5) hydrogen bond acceptors (nitrogen or oxygen atoms) ;
• a molecular weight under 500 g/mol ;
• a partition coefficient logP less than 5.

Ghose's rule
This rule establishes that  an orally active drug must have:

• partition coefficient logP in -0.4 to 5.6 range;
• molar refractivity from 40 to 130;
• molecular weight from 160 to 480;
• number of heavy atoms from 20 to 70.

The molecular refractivity is calculated according to the Ghose and Crippen method.

References:
Lipinski, C. A.; Lombardo, F.; Dominy, B. W.; Feeney, P. J.
"Experimental and computational approaches to estimate solubility and permeability in drug discovery and development settings"
AdV. Drug DeliVery ReV. 1997, 23,3-25.

Ghose, A. K.; Viswanadhan V. N.; Wendoloski, J.J.
"A Knowledge-Based Approach in Designing Combinatorial or Medicinal Chemistry Libraries for Drug Discovery. 1. A Qualitative and Quantitative Characterization of Known Drug Databases"
J. Comb. Chem. 1999, 1, 55 68.

It evaluates the electrostatic energy of the molecule in the current workspace. The default dielectric constant is 1 (vacuum).

You can choose between two prediction methods that use two different approaches to predict log P: the former, more accurate, is based on miLogP and requires to send the data to Molinspiration software and the latter, less accurate, is based on virtual log P and runs off-line. If you have to manage sensible data that you don't want to share on the Web, you should choose the second method.

The predictions miLogP-based exploit these two correlative equations:

log k_w^IAM.MG = -0.1405 + 0.4401 miLogP + 0.0536 HeavyAtoms - 0.0833 HLB_M - 0.0435 FlexDihedrals
n = 204   r² = 0.81   q² = 0.80   SE = 0.438   F = 213.92   P < 1.0 10-8   PC = 39.403

log k_w^IAM.DD2 = -2.3989 + 0.4936 miLogP + 0.4354 Vdiam - 0.0640 HLB_PSA - 0.0497 FlexDihedrals
n = 160   r² = 0.85   q² = 0.84   SE = 0.459   F = 212.94   P < 1.0 10-8   PC = 33.974

where:

The predictions virtual log P-based exploit these other correlative equations:

log k_w^IAM.MG = -0.3867 + 0.4159 VirtualLogP + 0.0741 HeavyAtoms - 0.0806 HLB_G - 0.0657 FlexDihedrals

n = 205   r² = 0.75   q² = 0.74   SE = 0.501   F = 151.79   P < 1.0 10-8   PC = 51.739

log k_w^IAM.DD2 = -3.0812 + 0.4809 VirtualLogP + 0.5464 Vdiam - 0.0765 HLB_PSA - 0.0829 FlexDihedrals

n = 161   r² = 0.80   q² = 0.79   SE = 0.523   F = 155.22   P < 1.0 10-8   PC = 44.319

where:

WARNING:
These two equations are valid only for neutral non-ionic molecules.

For more information about X-Score and XLOGP, visit http://www.sioc-ccbg.ac.cn/

It colors the molecule in the current workspace according the VMD color scheme.

Warning:

even if in the theory it's possible to manage a 2D database, adding the hydrogens by the bond order method and optimizing the structures, this procedure is not recommended because the distance geometry optimization is not performed. For this reason, a better choice is the conversion of the database from 2D to 3D (see the Database 2D to 3D.c script) and the resulting database can be used directly to predict the logP values.

• Start Weka and choose Explorer as application.
• In Process tab, click Open file... and select the ARFF input file. This file could require to be pre-processed to be usable (Use the Edit button).
• Go to Classify tab and choose the classifier (press Choose button in Classifier box). For example, select RandomForrest in trees.
• In the options of the classifier (click on the classifier parameters of Classifier box), set printClassifiers option to True to generate the right output with the tree models.
• Press Start button to generate the model.
• If the model is acceptable, save it by clicking with the right mouse button on Result list and choose Save result buffer. Put the file name adding .txt extension and press Save.

Decision tree conversion to C

• Run this program and select the output text file in Weka tree model.
• Optionally, you can choose also the ARFF file used to generate the model. Its data will be used to generate the code to check if the calculated attributes are include in the classification domain defined as range of the properties used to build the model.
• Set the output file name (Output C file).
• You can specify the name of the model if the proposed one is not ok (Model name field).
• Optionally, you can specify the labels printed as output for each class (Class labels field). The labels must be comma separated and their number must be the same of number of classes detected by the model.
• Select the source code type to generate. More in detail you can choose:
   
  • VEGA ZZ C-script: create a C-script for VEGA including the code to calculate the known molecular properties.
  • C source + header: generic code for classification. You can use it also not in VEGA.
  • Header only: same of above with the difference that all code is written in the header file.
• If you are building a VEGA ZZ C-script and you want to install it into VEGA ZZ environment, you can check Install VEGA ZZ C-Script and put the installation directory in Script directory field. If you leave this field empty, the C-Script will be installed in home directory. You can use the disk button to explore the directory tree and you must remember you cannot install scripts outside the directory tree of the scripts.
• Finally, click the Convert button.

When you choose VEGA ZZ C-script, the attribute names are analyzed and if are calculable by VEGA ZZ the right code is automatically added to the output, otherwise a warning message is shown because the resulting code will be incomplete and requires further implementations by the user.

1. If needed, adds the hydrogens by protein method.
2. If required, assigns the atom charges.

If the molecule has two dimensions only, the 2D to 3D conversion is performed as explained below:

3. Sends the molecule to AMMP.
4. Performs the Gauss-Siedel distance geometry optimization (15 steps).
5. Performs the steepest descent energy minimization (50 steps, toler = 1).
6. Performs the conjugate gradients energy minimization (3000 steps, toler = 0.01).
7. Sends the resulting structure to VEGA ZZ.

These steps are performed for both 2D and 3D structures:

8. Fixes the atom types, applying the AutoDock force field.
9. Removes the apolar hydrogens.
10. Saves the molecule in PDBQT format.

1. If needed, adds the hydrogens by protein method.
2. If required, assigns the atom charges.
3. Fixes the atom types, applying the AutoDock force field.
4. Removes the apolar hydrogens.
5. Saves the molecule in PDBQT format.
6. Runs AutoGrid4 to calculate the maps if the user confirms the operation.

The pre-defined docking box is set to explore the entire receptor, but if you want explore a specific protein region, you must select the atoms defining that region before to run the script.
The grid spacing is automatically adjusted if the number of grid points exceeds the 63 value because AutoGrid 4 and AutoDock 4 can't manage grid greater than 63x63x63 points.

In the graphic interface, some parameters can be set:

• Receptor
  File name of the target macromolecule (receptor) in Mol2 format.
• Ligand
  File name of the ligand to dock in Mol2 format.
• Output directory
  Directory in which the output files will be created. This field is automatically completed by selecting receptor and ligand.
• Flex. residues
  List of residues (space separated, in ResNum format) whose side chain will be considered flexible. By clicking the Get button, the field is automatically filled the residues that are active in the current workspace.
• Center
  X, Y, Z coordinates of the binding site center.
• Radius (Å)
  Radius in Å of the sphere including the binding site atoms. Clicking Calc. button, Center and Radius fields are automatically filled considering as binding site the visible atoms of the molecule shown in the current workspace.
• Clusters
  Number of solution clusters.
• RMSD
  Root Mean Square Deviation for the cluster analysis.
• Multimodel output
  Checking this gadget, it's possible to save all solutions in a whole multimodel file in Mol2 format.
• Atom scoring
  This checkbox allows the scoring values of each atom to be saved in the Mol2 output. The atomic charges are replaced by scoring values.
• Rigid ligand
  The ligand is kept rigid.
• Shape constraints
  In these fields, you can specify the molecule and the weight that is used for the volume overlap calculation (the more ligand atoms overlap, the better). For more information, read the PLANTS manual.
• Score
  Scoring function (chemplp, plp and plp95).
• Search
  Search mode: speed1 (highest reliability, slowest settings), speed2 (good reliability, twice as fast as speed1) and speed4 (modest reliability, four time fast as speed1).

By clicking Run button, the calculation starts and a window is shown in which it's possible to stop the run by clicking Abort button.

WARNING:

if you close VEGA ZZ, the PLANTS calculation is not stopped, but when it finishes, the scripts doesn't convert the output files to be read directly by Microsoft Excel.

PLANTS installation:

• Complete the registration form in download page at http://www.tcd.uni-konstanz.de/
• Download the PLANTS Win32 (minGW).
• Rename the file name to Plants.exe and copy it to ...\VEGA ZZ\Bin\Win32 directory, where ...\VEGA ZZ is the VEGA installation directory.
• Download mingwm10.dll and copy it to ...\VEGA ZZ\Bin\Win32 directory.

If you installed the 1.1 version built by Mingw32, it's strongly recommended to patch it by running Patch bin 1.1 script.

WARNING:

To run this script, you need the administrative rights, otherwiese it will be impossible to patch PLANTS. If User Account Control (UAC) is enabled, you must run VEGA ZZ as administrator. To do it, click the VEGA ZZ icon on the desktop with the right mouse button and select Run as administrator.

Rescore Plp.c

Rescore Plp95.c

You can give also a database as input including both receptors and ligands. The script try to detect automatically the ligand and when it's not possible, a requester is shown.

• Receptor
  File name of the target macromolecule (receptor) in PDBQT format.
• Ligand
  File name of the ligand to dock in PDBQT format.
• Output model
  File name of the ligand poses in PDBQT multimodel format. Remember that this file doesn't include the receptor structure.
• Log file
  Vina log file name.
• Center
  X, Y, Z coordinates of the binding site center.
• Size (Å)
  Dimensions in Å of the cube including the binding site.
• Exahaustiveness
  Exhaustiveness of the global search (roughly proportional to time).
• Binding modes
  Maximum number of binding modes to generate.

For more information about AutoDock Vina, click here.

1. If needed, adds the hydrogens by protein method.
2. If required, assigns the atom charges.

If the molecule has two dimensions only, the 2D to 3D conversion is performed as explained below:

3. Sends the molecule to AMMP.
4. Performs the gauss-Siedel distance geometry optimization (15 steps).
5. Performs the steepest descent energy minimization (50 steps, toler = 1).
6. Performs the conjugate gradients energy minimization (3000 steps, toler = 0.01).
7. Sends the resulting structure to VEGA ZZ.

These steps are performed for both 2D and 3D structures:

8. Fixes the atom types, applying the Vina force field.
9. Removes the apolar hydrogens.
10. Saves the molecule in PDBQT format.

1. If needed, adds the hydrogens by protein method.
2. If required, assigns the atom charges.
3. Fixes the atom types, applying the Vina force field.
4. Removes the apolar hydrogens.
5. Saves the molecule in PDBQT format.

• the receptor structure in PDBQT format. You can prepare it from any type of file using Receptor.c script;
• the database containing the ligands to screen. It must be in any format supported by VEGA ZZ (Microsoft Access, Merck MMD, Mol2 multimodel, ODBC data source, SDF file, SQLite and Zip archive). The database don't require to be prepared before the screening, because the script has the capability to detect the missing features and to fix them. In particular, it can add hydrogens using the best strategy, fix the atomic charges and to convert structures from 2D to 3D.

The graphic user interface of this script allows to setup the screening in easy way, changing the following parameters:

• Receptor
  File name of the target macromolecule (receptor) in PDBQT format.
• Energies
  Output file in localized CSV format containing the energy of the best pose for each ligand. The first column is the molecule progressive number (MolID), the second one is the molecule name (Name) and the third one is the Vina energy of the best pose (Energy).
• Ligand database
  Database of the ligands to screen.
• Output models
  File name of the Zip archive in which the poses in PDBQT multimodel format are stored. The script add a numerical suffix to file name that is incremented automatically every time in which the file size exceeds the limit of 2 Gb.
• Log file
  Log file name.
• Center
  X, Y, Z coordinates of the binding site center.
• Size (Å)
  Dimensions in Å of the cube including the binding site.
• Exhaustiveness
  Exhaustiveness of the global search (roughly proportional to time, default 8).
• Binding modes
  Maximum number of binding modes to generate (default 1).

Clicking Calc button, Center and Size fields are automatically completed using the atoms selected in the current workspace that will be considered as binding site.
Clicking Save cfg, you can save the current configuration that can be restored clicking Load cfg. The resulting .vcf file is not compatible with Vina, while that generated by Docking.c maintains the compatibility (see --config option of Vina).

About the restart
The restart procedure is automatically performed if the energy CSV file is found. You can choose to restart the calculation or to run it from the beginning by a requester window.

For more information about AutoDock Vina, click here.

Warning:

This script requires Fred2 installed on your PC. You can request/buy it at http://www.eyesopen.com/

• Name: Name of the ligand;
• Poses: Number of poses;
• Area: Area of the first pose;
• HyperArea: Shared area of the set of docking poses;
• DeltaArea: HyperArea - Area;
• RatioArea: HyperArea / Area.
• Volume: Volume of the ligand;
• HyperVolume: Shared volume of the set of docking poses;
• DeltaVolume: HyperVolume - Volume;
• RatioVolume: HyperVolume / Volume.

The multiple poses of the same molecule are detected by their names: they must share the same prefix followed by the underscore character ("_").

• the calculation requires at least two molecules in the workspace and if there are more than two molecules or the ligand is ambiguous, the script asks to specify the molecule ID of the ligand;
• the receptor is simplified keeping only the residues included in a spheroid of 3 Å around the ligand. The user can change this value in the script (VS_PROXIMITY constant);
• the complex geometry is optimized until the termination criteria GNORM is achieved. By default, this value is set to 10 (see VS_MOPACKEYS_MIN constant in the script);
• the heat of formation is evaluated for both ligand and receptor separated and complexed;
• the binding enthalpy is obtained by subtracting the two energies;
• the results are shown in VEGA ZZ console and copied to the clipboard, if requested by the user.

This script requires at least Mopac 2012 for Windows that is not included in VEGA ZZ package. For more information, see Installation of optional components.

To calculate the RPScore, the ligand must be a peptide/protein with the residue names indicated in the sequence.

• the database containing the ligands that were docked;
• the receptor file in any format supported by VEGA ZZ;
• the WarpEngine GRAMM output file (it have the .csv file extension);
• the output directory;
• the numbero of top ranked complexes that you want to extract.

By default, the script saves the complexes in IFF format and assigns CHARMM force field and Gasteiger-Marsili atom charges. These default parameters can be changed by editing the script code.

This script requires X-Score 1.2 or 1.3 for Windows that is not included in VEGA ZZ package.

For more information about X-Score, visit http://www.sioc-ccbg.ac.cn/

To install X-Score package in VEGA ZZ enviroment:

• Open the following Web site:
  http://www.sioc-ccbg.ac.cn/?p=42&software=xscore
• Complete the on-line registration form.
• Log-in with your credential and download X-Score package for Windows platform.
• Open the tar file by WinRAR or other suitable software able to unpack tar gizipped files.
• Extract xscore_win32.exe from xscore_win32\bin to ...\VEGA ZZ\Bin\Win32, where ...\VEGA ZZ is the VEGA ZZ installation path (usually C:\Program Files\VEGA ZZ).
• Rename xscore_win32.exe to xscore.exe.
• Extract parameter directory from xscore_win32 to ...\VEGA ZZ\Data directory. This last directory is hard to identify, because every Windows version creates it in a different place. To find it, open VEGA console from Start menu and type OpenDataDir.
• Rename parameter to Xscore.
• Now, you are ready to use X-Score script. If you want to use xscore.exe from command prompt, open VEGA console and use xs command, that is a shell script that fixes the environment variable required by X-Score.

This script performs the batch file format conversion of all molecules contained in a folder. Some parameters can be changed in the dialog window:

• Source dir.
  Name of the source directory in which the converting files are placed. Click Open button to show the directory requester.
• Destination dir.
  Name of the destination directory in which all converted files will be inserted. Click Open button to show the directory requester.
• Output format
  Use this list to select the output format.
• Compression
  Compression method (default none).
• Add hydrogens
  - None
  No hydrogens will be added.
  - Generic
  Generic organic geometry-based method.
  - Generic BO
  Generic organic bond order-based method.
  - Nucleic acid
  Nucleic acid geometry-based method.
  - Nuc. acid BO
  Nucleic acid bond order-based method.
  - Protein
  Protein geometry-based method.
  - Protein BO
  Protein bond order-based method.
• Include the connectivity
  If checked, the atom connectivity is included (if the file format supports it).
• Include the atom constraints
  If checked, the atom constraints are saved into the file (if the file format supports it).
• Normalize the coordinates
  If checked, the molecule is translated at the axis origin (0, 0, 0).
• Assign the Gasteiger/Marsili charges
  If checked, the Gasteiger - Marsili atom charges are assigned.
  
  Clicking Convert button, the conversion starts and clicking Close the dialog window is closed.

It shows the amino acid by selection and/or by chemical/physical properties.

It convert a Fasta into a text file. That's is useful to load it into Microsoft Excel.

It finds the histidine protonantion state (on NE2 or on ND1) using the CHARMM potential and swap the hydrogens (e.g. H-NE2 to H-ND1) according to the hydrogen bond energy. If the energy difference between the H-NE2 and H-ND1 tautomers is more than 2.0 Kcal/mol the hydrogen is placed on the nitrogen realizing a structure with lower hydrogen bonding energy. The starting structure must have the hydrogens.

This script moves the hydrogen atoms to the end of the atom list. In this way, you can obtain files split in two parts: the first one containing the heavy atoms and the second one, placed at the end, containing the hydrogens. As an example, that's useful to write mol2 files compatible with GOLD docking system.

• Electrostatic energy (Coulomb).
• Electrostatic energy with distant-dependent dielectric constant.
• R6-R12 Lennard-Johnes non-bond energy using the CHARMM and CVFF force fields.
• Hydrophobic interaction using the Broto-Moreau parameters with different distance functions (linear, square, cube and Ferm's function).

The results are automatically copied to the clipboard.

• Selection of the best independent variables by calculating the correspondent equation with a single regressor. Regressions with R² value less than 0.10 determine automatically the exclusion of the independent variable. If the number of found variables is less than the maximum number of regressors, the "desperate mode" is automatically enabled and 50% of the best variables are selected.
• Identification of collinear independent variables by calculating the Variance Inflation Factor (VIF) value for each regressor pair. Variable pairs with VIF > 5.0 are considered collinear and aren't not considered in the model calculation.
• Calculation of the models with a number of regressors from one to a user-defined value (default 3).
• For each model, a cross-validation procedure (leave-one-out) is performend and the prediction power is shown ad Q². If the number of observations is more than 200, the script asks to confirm the cross-validation.

The script requires a CSV file as input, that can be exported from your preferred spreadsheet software (e.g. Microsoft Excel) and generates an output file with the same prefix of the input followed by - regression.txt as name. The output file includes some information as the best independent variables, the collinear variable pairs, all regression models and the best regression models (three for each number of regressors).

• Trial
  Progressive number of the trial.
• Rsq
  Multiple correlation coefficient of the training set (r²).
• RsqAdj
  adjusted r² if the training set.
• PC
  Amemiya Prediction Criterion of the training set.
• P
  Probability of the training set.
• F
  Fisher F statistic for regression of the training set.
• StdDevOfErrs
  Standard deviation of errors (SE) of the training set.
• Test_MeanErr
  Mean error in prediction of the test set.
• Test_StdErrOfErrs
  Standard deviation of errors (SE) of the test set.
• Test_M
  Angular coefficient of the trend line of the chart of predicted vs. experimental activities.
• Test_B
  Intercept of the trend line of the chart of predicted vs. experimental activities.
• Test_Rsq
  r² of the chart of predicted vs. experimental activities.
• Test_PC
  Amemiya Prediction Criterion of the test set.
• Test_P
  Probability of the test set.
• Test_F
  Fisher F statistic for regression of the test set.
• Intercept
  Intercept of the regression equations.
• Coefficients of the independent variables
  The list of the coefficients for each regressor.

The output file includes also the mean (Mean) and the standard deviation (StdDev) of the previous columns and the labels of the columns selected as dependent (DepVar) and independent (InDepVar) variables.

  File name.....................: bestranking.csv
  Activity column...............: ACTIVITY
  Activity range................: 0.00 - 1.00
  Activity threshold............: 0.50
  Number of molecules...........: 2513
  Number of active molecules....: 38
  Max. minimization steps.......: 5000
  RMS to stop minimization......: 0.001
  Random sampling steps.........: 36
  Random selection probability..: 1.51 %

  Score = 1.0000 SCORE_0000 + 0.2309 SCORE_+000 - 0.5851 SCORE_0+00

   Top %  Mols   Act  Act %    EF
  =================================
    1.00    25     4  16.00  10.58
    2.00    50     6  12.00   7.94
    5.00   125    13  10.40   6.88
   10.00   251    18   7.17   4.74
   20.00   502    24   4.78   3.16

The coefficients of the equation are divided by the coefficient of the first term.

• ModID = identification number of the model that are ranked by ModelScore (from the best to the worst);
• NV = number of variables/scores used to build the model;
• Active_N = number of active molecules in the first N percentile;
• ActivePerc_N = percentage of the active molecules in the first N percentile;
• EF_N = enrichment factor in the first N percentile;
• Kurtosis = kurtosis of the histogram profile obtained by cluster analysis;
• Skewness = skewness of the histogram profile obtained by cluster analysis;
• ModelScore = score of the model (larger = better);
• Model = equation of the model with calculated coefficients (if you selected more than one score);
• ScoreMin = minimum score evaluated by the model;
• ScoreMax = maximum score evaluated by the model;
• Clustrer_N = percentage of active molecule in each cluster.

This script performs also the validation of the best models by building five pairs of training and external sets (with 70/30 % ratio) from the starting dataset. Training set is used to recalculate the models and external set to predict the activity. The results of this analysis are saved to prefix - valitadion.csv in which are present the same data as for the models obtained from the whole dataset with the exception of population of the clusters. The headers of the columns are named with ts and es prefix to identify respectively the training and the external sets.

• Output trajectory
  File name of the output trajectory. In the file requester, is it possible to select the output format (default Gromacs XTC).
• XTC comp. (1-6)
  Gromacs XTC compression ratio. It has a meaning only if the Gromacs XTC format is selected as output (default 3).
• Save the animation
  Check this gadget to save/render the animation (e.g. avi, mpeg, etc).
• Number of frames
  Number of frames to put in the trajectory (default 50).
• X rotation
  Rotation in degrees around the X axis (default 0). Negative values are allowed.
• Y rotation
  Rotation in degrees around the Y axis (default 360). Negative values are allowed.
• Z rotation
  Rotation in degrees around the Z axis (default 0). Negative values are allowed.
• Animate
  Push this button to create the animation trajectory.
   

• Input molecule
  File name of the input molecule. When you select a molecule using Open button, Input trajectory, Output trajectory and Output energy fields are automatically updated.
• Input trajectory
  File name of the input MD trajectory. When you select a new trajectory using Open button, Output trajectory and Output energy fields are automatically updated.
• Output trajectory
  File name of the output trajectory. When you select a new trajectory using Open button, Output energy field is automatically updated. In the file requester, you can select the output format.
• XTC comp. (1-6)
  Gromacs XTC compression ratio. It has a meaning only if Gromacs XTC format is selected as output.
• Output energy file
  File in which the energy values are stored (CSV format). It's available only if Mopac is selected as Minimization type.
• First frame
  Trajectory frame from which the quenching starts.
• Last
  Trajectory frame to which the quenching ends.
• Step
  Increment for the frame enumeration.
• Minimization type
  This field allows to select the minimization type: None (nothing is performed), AMMP (molecular mechanics method based on the conjugate gradients algorithms) and Mopac (semiempirical method).
• AMMP min. steps
  Number of minimization steps used by AMMP.
• AMMP toler
  I's the convergence criterion used by AMMP to stop the minimization.
• Mopac keywords
  In this field, you can put the keywords to control the Mopac calculation.
• Calculate
  Press this button to perform the quenching. If a parameter is incorrect or missing, an error message is shown.

• Input molecule
  File name of the input molecule. When you select a molecule using Open button, Input trajectory, Output trajectory and Output energy fields are automatically updated.
• Input trajectory
  File name of the input MD trajectory. When you select a new trajectory using Open button, Output trajectory and Output energy fields are automatically updated.
• Output energy
  Output energy file in CSV format (Comma Separated Values). Each column contains the following data: frame number, bond, angle, torsion, hybrid, non-bond and total energies.
• Output histogram
  Output histogram in CSV format.
• First frame
  Trajectory frame from which the quenching starts.
• Last
  Trajectory frame to which the quenching ends.
• Step
  Increment for the frame enumeration.
• Minimization type
  This field allows to select the minimization type: None (nothing is performed), AMMP (molecular mechanics method based on the conjugate gradients algorithms) and Mopac (semiempirical method).
• AMMP min. steps
  Number of minimization steps used by AMMP.
• AMMP toler
  I's the convergence criterion used by AMMP to stop the minimization.
• Mopac keywords
  In this field, you can put the keywords to control the Mopac calculation.
• Calculate
  Press this button to perform the quenching. If a parameter is incorrect or missing, an error message is shown.

• Input traj.
  File name of the input trajectory. Clicking Open button, the file requester is shown.
• Output traj.
  File name of the output trajectory.
• Output format
  File format of the output trajectory.
• Compression
  Compression level. It has an effect only if XTC format is selected.
• Append if the file exists
  If it's checked and the output trajectory exists, the converted frames are appended.
• Consider selected atoms only
  If it's checked, the active (visible) atoms only are saved into the new trajectory.
• Swap endian
  If it's checked, the endian of the converted trajectory is swapped. This function has an effect only if the DCD format is selected.

Click Go ! button to start the conversion and Cancel button to close the window.

This script performs the Ramachandran analysis for each trajectory frame. Before running it, you must open a trajectory file. For each frame, the Phi and Psi backbone torsion angles are measured and evaluated if they are inside or outside the Ramachandran permission areas. For each frame is calculated the percentage referred to the total number of the residues and these values are visualized in a plot. This calculation is useful to highlight the secondary structure evolution during a MD simulation. If the percentage of the residues (Phi and Psi values) inside the permission areas is decreasing during the simulation, it means that the secondary structure evolves to a worse situation. Vice versa, if the percentage is growing, the secondary structure is improving.

It eemoves all water molecules from a trajectory converting it into a PDB multimodel file. This script is obsolete and it's maintained as example only. The same function is now implemented in VEGA ZZ without external scripts.