Detect Binding Modes Panel

Cluster hypotheses that do not match all actives to detect possible different binding modes.

To open this panel, click the Tasks button and browse to Ligand-Based Virtual Screening → Detect Binding Modes or click the Workflow Action Menu button for a hypothesis group and choose Detect Binding Modes.

  • Using
  • Features
  • Additional Resources

Using the Detect Binding Modes Panel

This panel enables you to detect possible different binding modes when the hypotheses generated from a set of actives do not match all the actives. This can be the case when you are creating hypotheses from multiple ligands in the Develop Pharmacophore Model Panel.

The detection is done by performing hierarchical clustering on both the actives and the hypotheses to produce a specified number of clusters of each. The actives are clustered by the hypotheses they match and the hypotheses are clustered by the actives that match them. The results are presented in a table in which the rows (hypotheses) and columns (ligands) are sorted by cluster membership, and colored according to whether the ligand matches the hypothesis or not. The cluster boundaries are marked in the table with heavier lines to produce blocks. A binding mode is indicated by a block that is almost all blue (the color for hits), and the other blocks in the same row or column of blocks are mostly red (the color for misses).

You can change the number of clusters to consider by setting the value in the Proposed number of binding modes text box. This allows you to explore the possible binding modes.

Detect Binding Modes Panel Features

Load Results button

Load a file containing multiple hypotheses for the same system, as generated from the Develop Pharmacophore Model Panel. Opens a file selector so you can locate and select the file.

Proposed number of binding modes text box

Specify the number of binding modes that you think might be present. This is the number of clusters to use in hierarchical clustering. If you change the number, the clustering is repeated with the new number, and the table is updated.

Ligand-hypothesis match table

Show the results of the clustering in graphical form. The columns are the active ligands; the rows are the hypotheses. The rows and columns are sorted according to the clustering, and the cells are colored by whether the active matches the hypothesis. Heavy lines are drawn to delineate the clusters (binding modes).

When you select a cell for a match (blue) in the table, the ligand and the hypothesis are displayed in the Workspace. If you select multiple ligand cells for the same hypothesis, they are superimposed in the Workspace. If you select cells for multiple hypotheses, each hypothesis is displayed in a separate tile with the selected ligands.

Reset button

Clear the data in the panel and reset the settings to their defaults.

Related Topics