Section 2.4.9. MIF query applications

A MIF is suitable for interrogating databases because data items are permitted to have a single value, or a `sequence' of alternative values. This latter option is designated by the dictionary attribute _type_conditions which, for MIF applications, is set to `sequenced data' (via the code seq). This permits a value string to contain alternative `values' satisfying the following constructs: (a) the value string v1, v2, v3 signals that a data item must have the value v1 or v2 or v3, and (b) the value string v1:v2 signals that a data item must have a value in the range v1 to v2. Combinations of these constructions are permitted. All values must comply with the requirements defined by the attributes _enumeration and _enumeration_range.

An example of a substructural query in a MIF is shown in Fig. 2.4.9.1 for a conjugated ketone or thioketone fragment. Points of permitted variability of atom properties occur at atom 1, an sp³ carbon atom that must have at least one attached hydrogen atom, and at atom 5, which can be S or O. The conjugated multiple C—C bond (3–4) is defined to be either localized double, delocalized double or aromatic using CCDC bonding conventions. Query coding of this type should be readily generated from most graphical 2D search interfaces or be readable directly by a variety of 2D substructure search programs.

Figure 2.4.9.1 | top | pdf | Query substructure for conjugated ketones or thioketones. Atom C1 is sp3 hybridized (total number of attached hydrogen and non-hydrogen atoms = 4) and carries at least one hydrogen atom. Bond C3—C4 may be localized double (D), aromatic (A) or delocalized double (E) in CCDC conventions.