Applications - Design and Use of Relational Databases in Chemistry - page 159

Table 13.1 Sample of 10 Synonyms
in pubchem for inosotol 1-phosphate
Name
1-L-MYO-INOSITOL-1-P
1D-myo-inositol 3-monophosphate
1D-myo-inositol 3-phosphate
1L-myo-inositol 1-phosphate
D-myo-inositol 3-phosphate
L-myo-inositol 1-phosphate
inositol 1-phosphate
myo-inositol 1-monophosphate
myo-inositol 1-phosphate
about structures in the database. Most properties, such as experimental
values may have multiple values depending on how they were measured.
Experimental values belong in a separate table where information about
those values can be stored as well as the values themselves. For example,
the date of the measurement, uncertainties, methods, as well as multiple
values might be stored. When storing multiple values for any one struc-
ture, each value is stored as a separate row in the table with the same
structure id . The reasons for doing this were discussed in Chapter 2 in
the section on normal forms. It is possible to create additional tables in the
same schema, but this is not necessary. It might be useful to create another
schema for each type of experimental data, for example for each assay or
project. The way to store experimental data is discussed in a later section
of this chapter.
This approach for creating a registry of compounds might be
expanded in many ways. One important thing to consider is how tau-
tomers are handled. For some tautomers, a set of rules can be devised.
In principle, all tautomeric forms of any compound are valid and detect-
able chemical entities. But for some compounds, one tautomeric form is so
dominant, or the other forms are of so little interest, that one form can be
chosen as the standard tautomer. For example, all nitro groups could be
stored as charge-separated, for example C[N+](=O)[O−] or as CN(=O)=O.
Unless your projects are concerned with the very topic of tautomerization
in nitro groups, this is usually an arbitrary choice. It is usually helpful
to enforce a standardization rule to create a unique set of structures that
can be accurately searched. Rather than forcing users to remember the
rules, a table of transformations is used to change structures entered with
the “wrong” form into the standard form. This could be done using the
transformation and reaction functions discussed in Chapter 9. In order
to accomplish this, the add _ new _ structure trigger function shown
above could be expanded as follows.