This component allows users to review and optionally adjust the clustering solution produced by the Molecular Series Identification component. It is only relevant when a clustering-based series identification method has been selected (Murcko's scaffold cluster or Molecular cluster).
Although the optimal number of clusters is automatically determined by the Silhouette method and used by default, this component provides an Interactive View / Web Portal page (i.e. Clustering review) where users can evaluate the clustering solution and manually adjust the number of clusters if needed. The evaluation is facilitated by the inspection of the furthest objects — the two most distant and hence least similar scaffolds or compounds within each cluster — which provide an intuitive measure of cluster cohesion.
Workflow position
This component must be preceded by the Molecular Series Identification component and should be followed by the Molecular Series Selection component. When a non-clustering series identification method is selected, this component will result in an empty Interactive View / Web Portal page.
Inputs
- First input — must be connected to the first output of the Molecular Series Identification component
- Second input — must be connected to the distance matrix (2nd output) of the Molecular Series Identification component
- Third input — must be connected to the clustering model (3rd output) of the Molecular Series Identification component
Although the optimal number of clusters is automatically determined by the Silhouette method and used by default, this component provides an Interactive View / Web Portal page (i.e. Clustering review) where users can evaluate the clustering solution and manually adjust the number of clusters if needed. The evaluation is facilitated by the inspection of the furthest objects — the two most distant and hence least similar scaffolds or compounds within each cluster — which provide an intuitive measure of cluster cohesion.
Workflow position
This component must be preceded by the Molecular Series Identification component and should be followed by the Molecular Series Selection component. When a non-clustering series identification method is selected, this component will result in an empty Interactive View / Web Portal page.
Inputs
- First input — must be connected to the first output of the Molecular Series Identification component
- Second input — must be connected to the distance matrix (2nd output) of the Molecular Series Identification component
- Third input — must be connected to the clustering model (3rd output) of the Molecular Series Identification component
