Query substructure

Current chemical information systems offer three principal types of search facility. Structure search involves the search of a file of compounds for the presence or absence of a specified query compound, for example, to retrieve physicochemical data associated with a particular substance. Substructure search involves the search of a file of compounds for all molecules containing some specified query substructure of interest. Finally, similarity search involves the search of a file of compounds for those molecules that are most similar to an input query molecule, using some quantitative definition of structural similarity. 

Figure 8.2 Example of a 2D substructure search. The search is for the diphenyl ether query substructure at the top of the figure, below which are shown five of the hits resulting from a search of the National Cancer Institute database of molecules that have been tested in the US government anticancer program (see URL http //dtp.nci. nih.gov/). This database is also used for the search outputs shown in Figures 8.3 and 8.4.

Substructure searching involves retrieval of all the compounds in u hie containing some specified portion of a chemical structure, irrespective of Ihe rest of the molecule in w hich the query substructure occurs. 

Logic in Query Features. Using AND, OR, and NOT as modifiers on the application of query features. For example, one could run a search to select structures that contain "halogen and not primary or secondary amine, or not halogen and any amine." The logic can be a part of the query substructure, as with Markush queries, or it can be part of the SELECT statement. 

Substructure Search. Application of subgraph isomorphism" search to chemical structures. This consists of finding a particular arrangement of atoms and bonds as they are embedded in a chemical structure. The arrangement being searched for is termed the query substructure, the structures being searched are termed the candidates, and any particular structure in that set is termed a target structure. If the query substructure is... 

To perform a substructure search, both the query substructure and the molecules are represented by their Compressed Feature Matrices based on the same set of structural features. 

The fragments that have been chosen to act as screens are listed in a fragment coding dictionary. When a query or a new molecule is to be processed, the corresponding connection table is analyzed to identify those screens from the coding dictionary that are present in the structure. A database structure or query substructure is then represented by a fixed-length bit string... 

Substructure searching is the process of searching a file of chemical structure representations, not to find a particular compound, but to find any and all compounds which contain the same specified query substructure. 

Figure 1 illustrates the mapping between the nodes of a query substructure and a file structure in this case there are two possible correspondents for the nitrogen atom in the query - this is an example of a multiple isomorphism. 

Figure 1. Mapping between the nodes of a query substructure and those of a file structure.

Figure 6. A query substructure with free sites marked, and two possible answers (discussion in text).

Molecular similarity seemed to be ideally suited for the task, except for one important thing. Unlike the certainty of substructure searches that identified compounds that either did or did not contain the query substructure, similarity is a subjective notion and thus is more difficult to exploit since there is no right answer. For example, in similarity-based searching, knowing a compound s structure is not sufficient the key concern is what molecular features do the two molecules have in common. This begs a number of questions. What molecular features should be considered How is their relative importance assessed How is this information... 

Screens traditionally denote the presence or absence of predefined atom-, bond-, or ring-centered substructural fragments. However, one may also use subgraphs of the molecules to generate a set of molecular fingerprints to use as the screen. The screen search checks each structure for those screens present in the query substructure. For maximum effeaiveness the fragments included should occur independently and with equal frequency in the database. " ... 

Substructure searching may not identify molecules with minor deviations from the query substructure. To solve this deficiency, similarity-based searching techniques have recently been devised. They rank all the compounds in a... 

The basis of ALADDIN is the Daylight Chemical Information Systems software, particularly GENIE, a substructure specification language. When GENIE finds a query substructure in an input SMILES structure, it can return to the user those atoms in the structure that correspond to those hit. Since in a MENTHOR database the coordinates of the atoms are stored in the order in which they occur in the SMILES for that molecule, the coordinates of the atoms of interest are thereby identified. Thus, our geometric objects are established from this set of atoms, and geometric tests are performed on them. Steric tests are performed on molecules that meet the geometric criteria. 

Substructure searching, whether in 2D or in 3D, provides an invaluable tool for accessing databases of chemical structures. It does, however, have several limitations that are inherent in the retrieval criterion that is being used, which is that a database record must contain the entire query substructure in precisely the form that it has been specified by the user (subjea to any variations that have been specified, e.g., a distance tolerance or a range of acceptable substituent types for a specific position on a ring system). 

Secondly, there is very little control over the size of the output that is produced by a particular query substructure. Without a detailed knowledge of the contents of the file, the searcher will be unable to predict a priori how many database structures will satisfy the structural constraints defined by a given query. Even in the case of a 2D substructure search of an in-house file, the specification of a common ring system or of several possible substituents at a particular location (or locations) can result in the retrieval of several thousands of structures (unless it is also possible to specify other constraints such as... 

The query substructure for reagent searching (see Note 10) Enter the query into a text file using an editor like vi ... 

Please note that the implicit assumption was made that the user wanted to find sugars with the query substructure relative to the standard sugar orientation for this database. If the user really wanted the query substructure regardless of the orientation of the substructure relative to the rest of the molecule, the search would have to be repeated a second time with the query substructure rotated 180 perpendicular to the plane of the screen. 

The compounds in this database retrieved as matches to the query substructure are n-gulose, n-glucose, n-xylose, 4-p-n-glucopyranosyl-p-n-glucopyranose, and n-sucrose. 

---