BCUT descriptor

The BCUT descriptors (Buiden-CAS-University of Texas eigenvalues) [52], are commonly used, are eigenvalue-based, and include 3D information also. 

Hopfinger et al. [53, 54] have constructed 3D-QSAR models with the 4D-QSAR analysis formahsm. This formalism allows both conformational flexibility and freedom of alignment by ensemble averaging, i.e., the fourth dimension is the dimension of ensemble sampling. The 4D-QSAR analysis can be seen as the evolution of Molecular Shape Analysis [55, 56]. 

In 4D-QSAR, a grid is used to determine the regions in 3D space responsible for binding. Nevertheless, neither a probe nor interaction energy is used. 

---

Z eb index, Wiener index. Balaban J index, connectivity indices chi (x), kappa (k) shape indices, molecular walk counts, BCUT descriptors, 2D autocorrelation vector... 

Before the comparative molecular field analysis (CoMFA), BCUT descriptors, 4D-QSAR, and HYBOT descriptors arc discussed in more detail, some further descriptors are listed briefly. 

Dissimilarity and clustering methods only describe the compounds that are in the input set voids in diversity space are not obvious, and if compounds are added then the set must be re-analyzed. Cell-based partitioning methods address these problems by dividing descriptor space into cells, and then populating those cells with compounds [67, 68]. The library is chosen to contain representatives from each cell. The use of a partition-based method with BCUT descriptors [69] to design an NMR screening library has recently been described [70]. 

A significant percentage of any compound library will inevitably fall into small clusters unsuitable to rigorous statistical evaluation. These must be considered separately - in our case, using diversity analysis with BCUT descriptors [39] to supplement the list derived from clustering. Throughout this process, we use visualization to assess data quality, identify potential problems such as edge effects, and check trends and patterns. 

In the technique of post hoc design, a set of descriptors are built up by examination of a set of compounds active at a particular receptor family or sub-class. Normally, the set of drugs would be from a commercial database such as MDDR or the Merck Index, etc. and the descriptors would usually be substructural fragment or key based. One example would be the GPCR-PA+ sub-class referred to above, where BCUT descriptors have been used to aid the design of a focused library of aroimd 2000 compoimds based on 8 scaffolds. Libraries have also been constructed based on peptidomimetic principles as well as on the concepts of privileged structures. ... 

Fig. 2. Example of rough activity landscape. This figure shows the activity landscape for a series of related antibacterial compounds plotted in using the 2D BCUT descriptors to arrange the compounds. (A) Shows how the compounds are arrayed in a 2D representation of the chemistry space with the height of the marker being proportional to the minimum inhibitor concentration of the compounds [the smaller the minimum inhibitory concentration (MIC) the more potent the compound]. (B) This second panel presents the upper figure as a 2D figure to enhance the sharp cutoff between active and inactive compounds, emphasizing the point that activity landscapes are rarely smooth continuous functions.

Similar compounds were selected for further testing using BCUT descriptors and Euclidean distance to identify the untested compounds closest to the initial hit (24). 

Pirard, B. and Pickett, S. D. (2000) Classification of kinase inhibitors using BCUT descriptors. J. Chem. Inf. Comput. Sci. 40, 1431-1440. 

We use assay data from a National Cancer Institute HIV/AIDS database in our study (http //dtp,nci,nih.gov/docs/aids/aids data.html). As descriptors, we apply a set of six BCUT descriptors and a set of 46 constitutional descriptors computed by the Dragon software. These descriptors could be computed for 29,374 of the compounds in the database. The assay classifies each compound as confirmed inactive (Cl), moderately active (CM), or confirmed active (CA). We treat the data as a binary classification problem with two classes inactive (Cl) and active (CM or CA). According to this classification, 542 (about 1.8%) of the compounds are active. 

The UCC and clustering methods require a descriptor set—BCUT or constitutional descriptors. As our implementation of UCC requires continuous descriptors, the 46 constitutional descriptors, which include discrete counts, were also reduced to either the first 6 or the first 20 principal components (PCs). Thus, the UCC algorithm was applied to the BCUT descriptors and either 6 or 20 PCs from the constitutional descriptors. In addition to these three sets, clustering was also applied to the 46 raw constitutional descriptors. The random design requires no descriptors. 

Reported is the number of active compounds found by CBA when 100 or 200 top scoring compounds are selected from various designs, each replicated twice. In all cases, CBA uses the six BCUT descriptors for analysis. 

Table 4 presents the results for CBA. Four designs from the training data are considered, but only the BCUT descriptors are used here for analysis. CBA prioritizes the test compounds, and those with the highest scores are selected ... 

Fig. 4. (see facing page) Example tree. An RP tree was built using compounds selected by a space-filling design BCUT descriptors were used for the analysis. 

We computed an analysis of variance over Table 1, followed by tests of specific effects. Two results were statistically significant. Random selection of compounds was better than cluster or space-filling selection. BCUT descriptors were better for analysis than either of the principal component descriptor sets. 

The receptor relevance of BCUT descriptors has inspired several groups to apply them in conjunction with other methods. Beno and Mason reported the use of simulated annealing to optimize library design using BCUT chemistry space and four-point pharmacophores concurrently (33) and the use of chemistry spaces in conjunction with property profiles (52). The application of such composite methods to target class library design is readily apparent. Pirard and Pickett reported the application of the chemometric method, partial least squares discriminant analysis, with BCUT descriptors to successfully classify ATP-site-directed kinase inhibitors active against five different protein kinases... 

Burden, CAS, and University of Texas (BCUT) descriptors are well suited and widely used to describe diversity of a chemical population in a low dimensional Euclidian space and they allow for fast cell-based diversity selection algorithms (Pearlman and Smith, 1998). The DiverseSolutions... 

---