Building Multiple Feature Tree Models

In order to build an MTree model from more than two Feature Trees, the dynamic match-search algorithm can be applied in a hierarchical manner. We developed two efficient heuristics for this task  

MTree models constructed in this way can be used for screening purposes. By merging the information of the underlying trees into an MTree model, virtual screening can be done by simple pairwise comparisons. Hence we do not have to compare the ranks of several virtual screening runs comparing individual query molecules with each database molecule. 

Another advantage is that the matches in a model can be weighted by the local similarity of the corresponding subtrees in those matches, thereby pronouncing certain elements of the model. An exemplary application is provided in Section 4.6. 

---