ADAPT substructure descriptors

One of the earliest in silico models of human bioavailability was reported by Hirono and coworkers [26]. This study employed a set of 188 compounds that were classified as low (<50%), medium (50-89%), or well (>90%) absorbed and used a classification routine, fuzzy adaptive least squares, to generate discriminant functions. The molecules were described by their physicochemical properties and substructural descriptors, which meant that functional groups or substructures that enhanced bioavailability (e.g., saturated carbon atoms in side chains) or reduced it (e.g., aliphatic hydroxyl groups) could be identified. The performance varied between the three classes with the lowest success for the well-absorbed compounds, perhaps the most important group of the three. 

Figure 5- Substructures utilized by ADAPT ( ) to generate environment descriptors (from Ref. T).

The following concepts deal with graph and molecule substructures. These are necessary for introducing molecular descriptors, which are relevant for the applications described in later chapters. Our notation is adapted to the notation commonly used in chemistry, and is slightly different from that in graph theory. 

It is not trivial to define or select substructures that should be considered for this purpose. For the STIRS system, considerable effort has gone into the search for substructures that can be successfully classified by the implemented spectral similarity search. The Mass-Lib system uses a predefined set of 180 binary molecular descriptors to characterize the similarity of structures. In most investigations a more or less arbitrary set of substructures, functional groups or more general structural properties (compound classes) has been considered. Self-adapting methods that automatically analyse the molecular structures in the hitlist (for instance by searching for frequent and large substructures) have not been used up to now in MS. 

---