An Efficient Algorithm for Automating Classification of Chemical Reactions into Classes in Ugi’s Reaction Scheme

Graph Theoretic Representation of Molecules and Reactions

Chemical structures are usually stored in a computer as molecular graphs (Beck et al., 1969). A graph is an abstract structure that contains nodes connected by edges. One example of graph that corresponds to chemical structure of ethanal is shown in Figure 1. In such molecular graphs the nodes correspond to the atoms and the edges to the bonds. Such graphs can also be represented as matrices. Their major advantage is that the calculation of paths and cycles can be performed easily by well-known matrix operations. The matrix of a structure with n atoms consists of an array of n × n entries. A molecule with its different atoms and bond types can be represented in matrix form in different ways depending on what kind of entries are chosen for the atoms and bonds. Thus, a variety of matrix representations for chemical compounds are in use (Gasteiger & Engel, 2003): adjacency (Ivanciuc, 2000; Markovic et al., 2001), distance (Ivanciuc & Ivanciuc, 1999; Mihalic et al., 1992), incidence, bond, and bond-electron matrices. Among these representations bond-electron matrix representation provides all information given by bond matrix and additionally number of free electrons of an atom. In this work we have used BE-matrix, the representation of which is explained in the following.

Figure 1.

Ethanal can be represented in graph theory as a labelled graph