Reaction databases selected

The Theilheimer reaction database is distributed by MDL Information Systems, Inc., San Leandro, CA, USA. It contains 46 785 selected reactions from the years between 1946 and 1980. 

The values of rate constants of the reaction of selected phenoxyls with ionol calculated by the IPM method are given in Table 18.4. Experimental data on reactions of phenoxyls with phenols, hydroperoxides, and hydrocarbons can be found in Database [41] and those calculated by the IPM method values in the Handbook of Antioxidants [4]. 

ChemRXN 9.0 - Organic reaction databases include ChemSelect from Infochem GmbH and a selection from ISI s ChemPrep, for a total of over 29,000 reactions. 

The next step is to design a set of reactions to synthesize the compounds. One or more reaction databases can be searched to find whether any reactions give the desired structures as products or give structures that are similar to the desired ones. The chemist may also use reaction similarity searching (73)and searching across reaction schemes (e.g., if A + B c + D and C + E F + G a reaction scheme search wiU find the query A — F) (74). Once a reaction is found, the chemist needs to decide what reagents to use in the synthesis and where to obtain them. The selection of reagents will usually be based on a combination of physicochemical property considerations (i.e., QSAR and diversity), tempered by... 

Figure 9.21. Web client for an application that searches a relational reaction database. SQL statements are used to select structures and reactions that satisfy the search query.

Various new ways are described for the operator to make selections from the output. A program under development is described for interfacing SYNGEN with external reaction databases, in order to seek literature precedent for the generated reactions. 

The main goal for the software system under development is to make available in machine-readable form all the information contained in Cheminform. In this way different kinds of indexes may be produced as well as data deaUng with selected parts of chemistry for specialised information services, either printed or on computer-readable media. But the most important fact is that the information in Cheminform contains all the data needed for a reaction database. 

In May, 1989, Wipke and Vladutz presented a paper stud)dng the relationship and relevance of citations to a select group of reactions. They stated that all currently available reaction databases are selective in coverage and thus do not comprehensively include all related reactions. They also determined that 90-95% of the citing papers were relevant to the original reactions. This study is an extension of that work and with the objective scope being to determine the proportion of related reactions which might be found using a cited reference search approach. 

An analysis of the ORAC CORE reaction database of over 50,000 reactions shows that approximately 15% of the reactions are linked to other reactions via implicit links. We estimate that there are about 25,000 implicit links between reactions in the ORAC CORE database. However, implicit links between reactions in different selective databases are much less common due to the variation in the type of chemistry and examples used. The distribution of links between reactions is very dependent on the reactant and product structures. Common reactants, such as benzaldehyde, may account for a relatively high percentage of implicit links in a reaction database. 

Although reaction databases were relative latecomers in the electronic chemical information market, there is a relatively large, and still growing, number of databases of organic reactions. For the purposes of evaluating and selecting databases appropriate for the reaction query at hand, a categorization is useful ... 

Typical examples of comprehensive databases are CAS-REACT and CrossFire plus Reactions. Typical selective databases are Theilheimer/JSM and, to a lesser extent (because they are larger), CCR and Cheminform RX, all of which contain reactions selected from the primary literature, and the subsets of ChemReact (see Section 3.2,8). 

This simple classification of reaction queries illustrates the need for both comprehensive and selective reaction (and sometimes even compound) databases. There is some disagreement among chemical information specialists about whether there will still be a need for selective reaction databases in the future, in view of the expected enhancement of retrieval and post-processing procedures in large reaction databases (see Sections 4.2, 4.11, and 5). 

As Cheminform RX only began in 1991 and CCR (started in 1986) is not publicly available via hosts at present, CRDS/DJSMONLINE are important as the only publicly available selective reaction database that extends significantly back in time. 

Another important feature is meta-information about content and coverage (time, sources, selection policies) of reaction databases, and possible changes in these over time. At present. 

Since many of these 2.5 million reactions of the InfoChem Reaction Database are variations of the same type of reaction, InfoChem has developed a sophisticated selection concept based on the identification of all the different individual reaction types included in this file. This so-called Classification Algorithm is also used by MDL Information Systems, San Leandro, CA. Reactions with identical reaction centers and neighboring atoms are considered to be of one reaction type. An analysis of the 2.5 million reactions led to the identification of over 390 000 different reaction types. In the ChemReact database each reaction type is represented by only one selected example reaction. ChemReact was further refined to create ChemSynth, ChemReactlOO, and ChemReact41. 

More elaborate scheme.s can he envisaged. Thus, a. self-organizing neural network as obtained by the classification of a set of chemical reactions as outlined in Section 3,5 can be interfaced with the EROS system to select the reaction that acmaliy occurs from among various reaction alternatives. In this way, knowledge extracted from rcaetion databases can be interfaced with a reaction prediction system,... 

The position of the ehosen strategic bond locates the reaction center. To derive the reaction siibstrncture, the user can select the number of bond, spheres around the strategic bond which should be included. The reaction substructure obtained is then n.scd as the query for a reaction substructure search in the database. Figure 10,3-42 illustrates the first and second bond spheres around a selected strategic bond of a retrosynthetic step. 

And last not least, we will have to see further improvements in the graphical user interfaces of software systems and the retrieval systems of databases in order to make software and databases more acceptable to the chemical community at large. Software and databases should speak the language a chemist is used to, with hand-drawn chemical structures and reaction equations, or even imderstand the spoken word - and only provide the desired information selectively, not buried in a phe of unnecessary output. 

---