CDD Vault Visualization provides a “Clustering” feature for Molecules loaded into Visualization. Whether you're sending data from your private, secure CDD Vault or importing a data file into the Visualization tool, you are now able to cluster Molecules using CDD Vault's cutting-edge Visualization tool.
Once your data is loaded into the CDD Vault Visualization tool and structures are analyzed, a “plot type” for clustering is available. Clicking the Cluster analysis tool will:
- give you an option to set a Tanimoto similarity threshold
- perform the clustering algorithm across all structures using RDKit’s ECFC-4 circular substructure counts
- add a new “Cluster ID” column to the data table
- Cluster ID is now available as a parameter which can be used in all plots
As an example, one could have a scatter plot of Inhibition scores versus a chosen chemical property, where the dots are colored by Cluster ID.