Tree-based partitioning

PHASE applys a novel, tree-based partitioning algorithm to exhaustively identify spatial arrangements of functional groups that are common and essential to the... 

The common pharmacophore hypothesis is the description of how the ligand binds to the receptor. And Phase follows a tree-based partitioning technique to find the common pharmacophores. Immediately after entering this step, we can see the... 

In chemoinformatics research, partitioning algorithms are applied in diversity analysis of large compound libraries, subset selection, or the search for molecules with specific activity (1-4). Widely used partitioning methods include cell-based partitioning in low-dimensional chemical spaces (1,3) and decision tree methods, in particular, recursive partitioning (RP) (5-7). Partitioning in low-dimensional chemical spaces is based on various dimension reduction methods (4,8) and often permits simplified three-dimensional representation of... 

In addition to cell-based partitioning, statistical partitioning methods are widely used for compound classification. One of the most popular approaches is recursive partitioning (Rusinko et al. 1999), a decision tree method, as illustrated in Figure 1.8. Recursive partitioning divides data sets along decision trees formed by sequences of molecular descriptors. At each node of the tree, a descriptor-based decision is made and the molecular data set is subdivided. For example, a chosen descriptor could simply detect the presence or absence of a structural fragment in a molecule. Alternatively, the... 

Another nonparametric regression method is CART (classification and regression trees). The basic concepts were outlined in Discriminant Analysis Section about tree-based classification. We remember from that chapter that CART is a recursive binary partition method based on a simple model constant for each region. If the residual sums of squares of responses is minimized. 

While recursive partitioning is highly interpretable, it also has a few undesirable characteristics. This method can produce predictive models that may not be scientifically interpretable, especially if the meaningful predictors are correlated with less meaningful pr ictors. Tree-based routines may select the less meaningful predictors for the partition if it meets the optimization criteria. Hence, the user should preprocess the data to identify and remove highly correlated predictors, keep scientifically relevant predictors in the model, and remove correlated, but less relevant predictors. 

Cluster analysis (CA) performs agglomerative hierarchical clustering of objects based on distance measures of dissimilarity or similarity. The hierarchy of clusters can be represented by a binary tree, called a dendrogram. A final partition, i.e. the cluster assignment of each object, is obtained by cutting the tree at a specified level [24],... 

A commonly used technique for analyzing drug discovery screening data from individuals is recursive partitioning (RP), more commonly known as trees (see, for example, Blower et al., 2002). In very recent times, efforts based on multiple trees (Svetniket al., 2003) have become the method of choice, despite the additional difficulties associated with them, because of their good predictive abilities. 

Fig 13 General hierarchies based on partitions of 6,7 and 8 respectively Top levels correspond to linear acenes, star trees and perfect graphs while bottom levels are those of the zigzag polyacenes and paths ... 

---