The navigation in the Ligand-Based Modeling Perspective is quite similar to the Library View , but with a different focus. The Ligand-Set Table (after importing the ligands) shows the list of molecules which serve as input for the pharmacophore generation process. The user can change ligand properties (e.g. type, activity, etc.), filter the table content, and select ligands to make them visible in the 3D (only if 3D coordinates are present), 2D, and Hierarchy View. Like in the Library View , navigating through available conformations is possible by using the Alternatives Switcher .

The ligands are colored differently to keep the overview. The Training-Set ligands are colored randomly and ligands from the Test-Set and Ignored Ligands are marked gray. The color of the ligands can be changed in the Hierarchy View . You can customize the type of a selected ligand by choosing the appropriate type in the Ligand-Set Table or flag the ligand in the Ligand-Set menu. It is also possible to change the activity of a ligand in this way.

In the Ligand-Set Table toolbar, buttons are provided for the main tasks in ligand-based pharmacophore modeling. From left to right: ligand conformations can be created by OMEGA ( Generate Conformations for Ligand-Set icon), the available ligands can be clustered according to a multi-conformational alignment score ( Clustering icon), and the automatic ligand-based pharmacophore creation ( Run Ligand-Based Pharmacophore Creation icon) can be started. If conformations are already available, the pharmacophore generator uses them directly for the pharmacophore creation, otherwise they are generated on the fly. Conformer generation, clustering and pharmacophore creation can be customized as described in the section called “Ligand-Based Modeling Settings” and the section called “Alignment Settings” .

When a ligand-based pharmacohore creation run is successful, the Ligand-Set Table is automatically updated and provides additional information such as feature patterns, number of conformations, and the Pharmacophore Fit of the ligands matched with the selected result pharmacophore. The first digit of the Pharmacohphore Fit score represents the number of matched features and the second one, the RMSD value. The Feature Pattern of a ligand illustrates the color-encoded features that were matched by the ligand. Thus you can see at first glance the distribution of the features and which of the resulting pharmacohores are matched best by the Ligand-Set (frequency of matched features).

The Results Table is positioned at the right-bottom corner of the Ligand-Based Modeling Perspective. After the ligand-based pharmacophore generation process has finished the results are listed in this table including the name and score. By default, the entries are sorted by score value in descending order, but the user can change both the sorting order and sorting criteria by clicking the appropriate column identifier.

Each entry of the Results Table represents a valid pharmacophore model (i.e. it consists of at least three features) and stores the state of all Training-Set molecules (i.e. alignment pose and active conformation). Selecting a resulting pharmacophore model updates the 3D View and 2D View to show the particular pharmacophore and aligned Training-Set molecules. Furthermore, score values and feature patterns in the Ligand-Set Table are updated for the currently selected pharmacophore.

The range of the score values shown in the Results Table depends on the scoring function used (see Ligand-Based Modeling Settings ). If you select the Pharmacophore Fit scoring function then the score value describes the number of matched features and RMSD. All other scoring functions produce normalized values that range from zero to one, where one is the optimum.

The generated ligand-based pharmacophores may be transfered into other perspectives, where you can export them in different file formats. Optionally, you can generate ligand-based pharmacophores by using the command line as well. For further information, please see Chapter 8, Ligand-Based Pharmacophore Design .

During the ligand-based pharmacophore modeling process, ligands are ranked by their number of conformations. Then, starting from the ligand with the lowest number of conformations, their pharmacophore models are aligned with the ones of the second ranked ligand. Ligand-based pharmacophores will be composed from these alignments and aligned with the pharmacophores of the next ranked ligand and so on. If you set a user-defined pharmacophore in advance, this pharmacophore is kept as the first pharmacophore to be aligned to the pharmacophores of the following ligand. The advantage of this functionality is that the features of the user-defined pharmacophore are more likely to survive this procedure.

You can use this functionality by selecting in the menu Ligand-Set > Set Pharmacophore Bias and load a pharmacophore of your choice. Alternatively, you can add a pharmacophore from another perspective to the Ligand-Based Modeling Perspective using the Copyboard Widget .

Although, the first functionality stamps the ligand-based pharmacophore modeling by a user-defined pharmacophore model, LigandScout allows you to provide additional customized feature definitions in addition to LigandScout's standard feature definitions (built-in types). These chemical features will be used temporarily for aligning the molecule, but will not be included in the resulting 3D pharmacophore. You can access the custom feature functionality directly in the Ligand-Set menu or via the settings dialog after clicking the Run Ligand-Based Pharmacophore Creation button in the Ligand-Based Modeling Perspective.

A custom feature is defined by several SMARTS patterns that specify the substructures where the feature shall (or shall not) be placed. For the definition of a new custom feature click on the New button under the available Features panel and specify an ID (abbreviation) and name for the feature. To enable/disable the feature in the pharmacophore generation process click the Active check box in the feature table.

To define or include SMARTS patterns for a feature click on the New button under the Include Patterns List. Declare valid SMARTS patterns and specify tolerance spheres for the feature. Smarts patterns that specify substructures where the particular feature shall not be placed (exclude patterns) are defined in the same way except that you do not need to specify a tolerance.

If you want to make changes to your custom feature definitions, you can edit them by clicking on the Edit button. To delete custom features as a whole or to delete only some include/exclude patterns of a feature, select the desired features or patterns and click on the appropriate Delete button. To apply your changes click the OK button on the bottom-right corner of the Edit Custom Features dialog. To reset any of your changes click on the Reset Changes button.

Use the custom feature functionality when you want to put emphasis on certain common substructures of the ligands in the alignment steps performed during the pharmacophore generation process. Note that although the custom features will be considered in the pharmacophore creation process (if enabled) they are not visible in the resulting 3D pharmacophores.