Chemical structures representation structural keys

Of course, it will be important to store the chemical structure itself. Clearly, a complete chemical structure cannot be represented using basic SQL data types, such as numeric or text. While a compound name (text) might be considered a good representation of molecular structure, there are better ways to represent molecular structure. In this chapter, a compound will simply be identified with a unique compound id serving as a foreign key to a more complete representation of structure in another table. The following chapters will show ways that chemical structures can be fully integrated into the tables of a database. This chapter concentrates on the proper use of chemical data and ways in which multiple relational tables can be used. 

Key to the implementation of the substructure searching system is a chemical structure representation known as the ROSDAL string. This is an ASCII character string that conveys sufficient information to convey fully a structure representation, including stereochemistry. A subset of a structure, that is to say, a substructure query, can also be represented by the ROSDAL string. Use of the ROSDAL string provides two immediate advantages for any new substructure... 

Figure 9.1 Schematic representation of some of the simple network architectures structurally characterised for metal-organic polymers (a) 2D honeycomb, (b) ID ladder, (c) 3D octahedral, (d) 3D hexagonal diamondoid, (e) 2D square grid, and (f) ID zigzag chain (reprinted from Section Key Reference The American Chemical Society).

There is a wide variety of descriptors available to describe molecules the molecular representation they encode is key to the measurement of diversity. Descriptors directly influence the metrics and algorithms used in the design or analysis, the nature of the chemical space in question, and the location of molecules within the chemical space. Therefore, it is important to select descriptors most appropriate to the problem at hand. Such a selection of the appropriate descriptors is nontrivial, the requirement of compound representation is that it contains enough information to incorporate structure and, in some investigations, function. 

The number of features combined in a vector-type representation is indicative of the dimensionality of the problem space. Low-dimensional representations, on the one hand, allow easy visualization but are most often not very discriminative. Highdimensional representations, on the other hand, such as those encoded in Daylight fingerprints [23], MACCS keys [24], or UNITY fingerprints [25], provide more detailed accounts on structural or chemical variations. However, this is achieved at the cost of visualization. Part of these high-dimensional representations describe specific local features of molecules, and because not all molecules in the data contain these features, gaps or zeros are introduced in the data representation. For certain data mining methods, this could be problematic. In many cases, dimensionality reduction procedures are applied to reduce the complexity of the representation. The reduction of the dimensionality is accomplished by means of 1) variable selection procedures, 2)... 

---