QUASAR Tutorial - Graphical user interface (GUI)

The parser panel lets you define all imporant options connected to the alignment file you want to use and the alignment parser that is used to read the file format of the specified alignment file. Additionally you can tell the quasar system if it should try to parse 3D information (for example the PDB files of the templates used) or if you don't want to provide this information. If you want to provide 3D information you have to select a pdb file directory as well as a pdb file extension. Below you can see a screenshot of the parser panel:

The parser panel lets you define the alignment parser that fits your alignment format, the alignment file that contains the alignments that should be ranked as well as the option if you want to provide 3D information (and if yes how to retrieve this information) or not.

• Alignment parser: please select an alignment parser from the drop down list. The description of the alignment parser will be shown if you click on the "View description" button. The description contains the alignment format parsed by the parser as well as an information about the protein dimension provided by the parser. By protein dimension we mean sequence information (1D), secondary structure information (2D) or 3D information that can be provided by using the components described below. The information about the dimension that is provided by the parser is important because of the scoring schemes used later on. Since they require a certain protein dimension or more protein dimensions to work (for example a Q3 score can only be calculated if 2D information is available) it is important for the validity of the configuration that all the information required by the scoring schemes is present. So please make sure that you provide all the information required to use the scoring schemes you want to use. If you want to save a configuration file or run Quasar with an incorrect configuration, Quasar will ask you to correct the errors.
• Alignment file: This field lets you specify the alignment file name and the directory where to find the alignment file. Just type the information into the text field or click on the "select" button to open a dialog that lets you choose the file. Please make sure that alignment file and selected parser fit together.
• Parse 3D information: By setting this flag (check the checkbox) you can determine if 3D information for the template sequences i.e. PDB files should be parsed by Quasar. The 3D information has to be available in form of PDB files while only ATOM records are parsed. For some Quasar functionalities e.g. the Benchmark mode 3D information if necessary.
• PDB directory: It might be best to give an example here: lets say your alignment file contains the names of the template pdb files in the first line as shown in the image above when using Parser 1. Your PDB directory should now contain those pdb atom files (only ATOM entries in pdb format are parsed out). If the pdb files are not directly contained in the directory Quasar tries two retrieve the files from subdirectories following the ASTRAL and PDB subdirectory system.
• PDB file extension: to complete the example described above you have to provide a extension for your pdb files. If this extension is already given together with the template pdb file name read by the parser, just leave this field empty. If only the name of the pdb file is given, please write the extension here for example pdb, ent....
• Parse CA atoms only: if PDB coordinates are parsed (see above) one can choose between parsing all coordinates (all atoms) or just parsing CA coordinates. Parsing all atoms might be useful for certain energy functions since the distances between the amino acid side chains can be calculated and not only CA distances are available. On the other hand parsing all coordinates takes more time and requires more main memory. MaxSub, Touch and APDB do only require CA information.

This panel lets you define the options connected with the scoring formula used to score the alignments. As soon as you decided which score conductor you want to use you have to configure this score conductor. As soon as you finished the score conductor configuration and the parser configuration, you can run Quasar from the graphical user interface (described below) and / or store the configuration file to save your configuration. Below you can see the scoring panel in detail:

The image shows the scoring panel. It allows you to select and configure the scoring formula used to score the alignments.

• Score conductor: First of all you have to select a score conductor from the drop down list that should be used to score the alignments. In most cases the score conductor requires additional configuration described below. In the text area below you can find a description of the score conductor.
• Score conductor configuration: After you have selected a conductor you have to configure this conductor (otherwise you won't be able to save or run the current configuration). So if you want to configure a score conductor, please click on the open configuration dialog button. You can always click on the button again if you want to edit an existing conductor configuration. Please notice that if you select another score conductor, you loose the current configuration.

Score conductor configuration dialog

If you have clicked on the open configuration dialog button in the scoring panel, you open the configuration dialog for the currently selected score conductor. This dialog can look quite different from conductor to conductor since every score conductor requires its own, specific configuration information. If you don't know how to use such a dialog, you should try to see the tooltips of the components that should give you an idea about the function or read the conductor description more carefully. Nevertheless we want to show you one example for such a dialog here (generated to configure the WeightedSumConductor) but please don't be confused if the dialog you get looks different!

Configuration dialog for the WeightedSumConductor score conductor

This score conductor allows you to add one or more scoring schemes. To add such a scoring scheme please click on the add button (you can also remove or edit the currently selected scoring scheme afterwards by clicking on the corresponding buttons). If you clicked on the add button another dialog pops up that lets you configure the scoring scheme. As soon as you have finished your configuration please press the Finish configuration button.

Scoring scheme configuration dialog

Again the configuration dialogs for scoring schemes are different from scheme to scheme and also different if they are used together with different score conductors. So again, don't be confused with this. Here follows an example for the configuration of a SOV measurement scoring scheme in the context of a WeightedSumConductor.

This dialog allows you to add a scoring scheme to the score conductor

In this case you have the possiblity to select a scoring scheme in the drop down selection box (you see the description of the scoring scheme in the text area below) and to specify the weight of the scoring scheme in the weighted sum as well as if the score should be normalized to a value between 0 and 1. As soon as you have finished your configuration please press the Finish configuration button.

If you are satisfied with the selected score conductor and its configuration, you can either save the current configuration by using the File menu -> Save configuration function or you can select the Run panel to run Quasar with the current configuration from the graphical user interface.

The run panel allows you to run the Quasar system with the current configuration which might be useful for example to test a current configuration.

This image shows the Quasar run panel

• Console: all the output of the quasar system runs that were started via the graphical user interface is shown in the console window. If an error happens, this error is printed out in red.
• Run Quasar: click on this button to run Quasar with the currently specified configuration. If there are any errors in the configuration (for example, a scoring scheme requires an information that is not provided by the parsers) the system informs you about the error and allows you to correct it.
• Clear Console: press this button to clear the console
• maximal VM memory: Since the Quasar system is implemented in Java it uses a so called virtual maschine to run. This detail is not important to you but if you want to use very large alignment files that contain many alignments the current main memory the Java VM is allowed to use might not be enough to run Quasar. In this case a error called OutOfMemoryException is shown and Quasar exits. If this is the case, you have two possiblities: use a smaller number of alignments or extend the amount of main memory the Java VM is allowed to use. You can do this by typing another value into the given field. Please do not allow the Java VM to use more main memory than you have on your maschine.

First of all this panel allows you to add a number of benchmark scores to the current Quasar configuration. This means that if you run the Quasar system with such a configuration file in benchmark mode Quasar will not only calculate the quality score of the alignments as definded by your formula but it will also calculated the selected "objective" alignment scores added to the list in this panel. Additionally Quasar will calculate the correlation coefficient between your quality score and every "objective" benchmark score which could give you an idea about how your scoring formula relates to a real, objective measurement like RMSD....
Second, this panel allows you to run Quasar in benchmark mode with the given configuration (scoring formula and benchmark scores). To run Quasar in benchmark mode properly, you should make sure that:

• PDB ATOM files for all queries and templates are available otherwise benchmarking does not make sense
• You've selected a parser that is a valid benchmark parser in terms that it parses the 3D information for the template and the query. An example for such a parser is the "Benchmark parser one".
• You've selected and configured one or more benchmark scores

The benchmark panel does support the selection of the benchmark scores and starts a Quasar run in benchmark mode but it does not test if the parser you've selected is correct and if all the pdb files are available!

This image shows the Quasar benchmark panel

• Console: all the output of the quasar system runs that were started via the graphical user interface is shown in the console window. If an error happens, this error is printed out in red.
• Run Quasar: click on this button to run Quasar with the currently specified configuration and selected benchmark scores in benchmark mode. If there are any errors in the configuration (for example, a scoring scheme requires an information that is not provided by the parsers) the system informs you about the error and allows you to correct it.
• Clear Console: press this button to clear the console
• maximal VM memory: Since the Quasar system is implemented in Java it uses a so called virtual maschine to run. This detail is not important to you but if you want to use very large alignment files that contain many alignments the current main memory the Java VM is allowed to use might not be enough to run Quasar. In this case a error called OutOfMemoryException is shown and Quasar exits. If this is the case, you have two possiblities: use a smaller number of alignments or extend the amount of main memory the Java VM is allowed to use. You can do this by typing another value into the given field. Please do not allow the Java VM to use more main memory than you have on your maschine.
• List: The list window contains all the currently selected benchmark scores and allows you to add, remove and edit existing scores.
• Add: Click on this button to add a new benchmark score to the system. After clicking on the button for example this window appears:
  
  This window allows you to select a benchmark score from the drop down list (description of the currently selected score in the text area below) and to configure the score (just like scoring schemes...) and to add it to the list by clicking on the "Add score to benchmark" - button.
• Remove: if you selected a benchmark score in the list by clicking on it you can remove this score by clicking on the remove button
• Edit: if you selected a benchmark score in the list by clicking on it you can edit this score by clicking on the edit button

The optimization panel allows you to optimize the weights of your current scoring function. To do this, just select an optimizer from the drop down list, provide some configuration information and run the optimization. Quasar will write a valid configuration file containing the optimized weights. You can reload this file into the GUI and run Quasar using this configuration.

This panel can be used to visualize the output of Quasar. Therefore just open a *.ranking file that has been generated by the Quasar system (in benchmark mode) and select the scores you want to plot against your standard of truth score.

If you do a right mouse click into the plot screen you have access to several plot options. You can also zoom in and out in the plot.