Descriptor field

Descriptors have to be found representing the structural features which are related to the target property. This is the most important step in QSPR, and the development of powerful descriptors is of central interest in this field. Descriptors can range from simple atom- or functional group counts to quantum chemical descriptors. They can be derived on the basis of the connectivity (topological or... 

Once the molecules are aligned, a molecular field is computed on a grid of points in space around the molecule. This field must provide a description of how each molecule will tend to bind in the active site. Field descriptors typically consist of a sum of one or more spatial properties, such as steric factors, van der Waals parameters, or the electrostatic potential. The choice of grid points will also affect the quality of the final results. 

Schuffenhauer A, Gillet VJ, Willett P. Similarity searching in files of 3D chemical structures analysis of the BIOSTER database using 2D fingerprints and molecular field descriptors. J Chem Inf Comput Sci 2000 40 295-307. 

Here, we introduce frequently used descriptors classified by the structure of the descriptor itself linear, treelike, and field descriptors. 

As the ComPharm overlay procedure is stochastic, it is typically repeated five times and the resulting average pharmacophore field intensities at the reference atom centers are used as ComPharm Field Descriptors . 

A QSAR approach allowed to select relevant ComPharm field descriptors is therefore expected to probe which of the features represented in the reference compound (typically chosen to be a potent inhibitor) must be conserved and which can be altered without... 

Upon inclusion of ComPharm field descriptors in the initial pool of molecular indices, four of these are selected by the GA-driven QSAR building procedure to enter the minimalist six-variable overlay-based QSAR model. These ComPharm key points are... 

Fig. 5.4 Comparative display of a catalyst-HipHop hypothesis and the pharmacophore field descriptors selected by the minimalist ComPharm overlay-based model. ComPharm key features are pinpointed by arrows, while HipHop feature spheres stand for hydrophobes (light blue) and hydrogen bond acceptors (green).

At identical variable number, the overlay-based model outperforms its overlay-independent equivalent training RMS = 0.712 (pICso units), training = 0.712, validation RMS = 0.698, training = 0.724. ComPharm field descriptors therefore appear to have a better predictive power than pairwise pharmacophore feature counts alone. 

One of the most attractive features of the CoMFA and CoMFA-like methods is that, because of the nature of molecular field descriptors, these approaches yield models that are relatively easy to interpret in chemical terms. Famous CoMFA contour plots, which are obtained as a result of any successful CoMFA study, tell chemists in rather plain terms how the change in the compounds size or charge distribution as a result of chemical modification correlate with the binding constant or activity. These observations may immediately suggest to a chemist possible ways to modify molecules to increase their potencies. However as demonstrated in the next section, these predictions should be taken with caution only after sufficient work has been done to prove the statistical significance and predictive ability of the models. 

It is clear that for an unsymmetrical data matrix that contains more variables (the field descriptors at each point of the grid for each probe used for calculation) than observables (the biological activity values), classical correlation analysis as multilinear regression analysis would fail. All 3D QSAR methods benefit from the development of PLS analysis, a statistical technique that aims to find the multidimensional direction in the X space that explains the maximum multidimensional variance direction in the F space. PLS is related to principal component analysis (PCA)." ° However, instead of finding the hyperplanes of maximum variance, it finds a linear model describing some predicted variables in terms of other observable variables and therefore can be used directly for prediction. Complexity reduction and data... 

Three-Dimensional Holographic Vector of Atom Interacting Field descriptors = 3D-... 

Three-Dimensional Vector of Atomic Interaction Field descriptors = 3D-VAIF descriptors > Threshold Toxicological Concern —> property filters (0 functional group filters)... 

I 3D-VAIF descriptors (= Three-Dimensional Vector of Atomic Interaction Field descriptors) These are vectorial descriptors derived by an approach similar to that of MEDV-13 descriptor and defined in terms of nonbonding interaction energies between pairs of atom types [Zhou, Zhou et al, 2006]. Five atom types are defined on the basis of the chemical element of the most occurring atoms in organic compounds these are (1) H (2) C (3) N or P (4) O or S (5) F, Cl, Br, or I. 

The retrieval of all 3D structures from a database considered to be similar to a given target structure is comparable with 2D similarity searching. 3D similarity searching raises the problem of conformational flexibility. Schuffenhauer et analysed the BIOSTER database by similarity search using 2D fingerprints and molecular field descriptors. A comprehensive overview on pharmacophore perception and 3D database searches is given in references 101 and 102. 

---