Descriptors that are not Structure-Based

The theme of this book has a strong emphasis on the structure of chemical compounds, its representation, and its correlation with properties. However, there are compounds whose structure is either unknown or ill-defincd. This is true, for example, for many polymers, particularly those that have been prepared from several components, or for many technical materials, such as glues, washing powder, etc. 

One way of rcprcsciitiiig such compounds is to characterize them by the components from which they were made, the relative ratios that were utilized, and the reaction conditions that were employed to make them. 

Properties that come foremost to mind to represent a compound are physical ones, because most of them can be measured easily and with high accuracy. Clearly, the more properties are used to characterize a compound, the better a model can be established for the prediction of the property of interest. Furthermore, one should select such properties which one knows or assumes to have a strong influence on the property that one wants to predict. 

A particularly good selection of physical properties may be spectra, because they are known to depend strongly on the chemical structure. In fact, different types of spectra carry different kinds of structural information, NMR spectra characterize individual carbon atoms in their molecular environment. They therefore correspond quite closely to fragment-based descriptors, as underlined by the success of approaches to predict NMR spectra by fragment codes (see Section 10.2.3). 

Infrared spectra are strongly dependent on the 3D structure of a compound, as reflected by the success of attempts to simulate infrared spectra from 3D structure representations (see Section 10.2.5). Infrared spectra should therefore be taken as representations of a chemical compound for modeling properties that are suspected to be dependent on the 3D structure of a compound NMR spectra also depend on 

---