Descriptions of reactions: Their logic and applications

Abstract

Abstract. Three systems of reaction description, from Ugi, Fujita, and Hendricksori. are examined and compared. They are shown to be essentially identical, and are rationalized here into a single simplified representation. The application of these systems to reaction retrieval, synthesis, and the invention of new reactions is also described.In its long history organic chemistry has derived no generally used systematic description for the variety of its reactions. Nevertheless, such systems for the logical organization of organic reactions are important for the retrieval of reactions in a systematic way from catalogs or databases, for rational design of organic synthesis routes, and for the invention of new reactions. This paper will compare several major systems that have been put forward and rationalize them in a simple single representation'. This is followed by an overview of ways in which systematic representations are central to the development of these three areas. A general system of description requires an initial basis of clear simple logic which is then applied all possible ways to generate all possible reactions. As with all such general systems, this application inevitably results in a very large number of final permutations which can quickly become unwieldy, even though the initial system basis may be quite simple'. Hence it is important to describe the system in such a way as to recognize its permutation possibilities while at the same time maintaining a chemist's simple perspective on the whole through clear visualization.

Chemistry, Computers, and the Mind

“…published) reactions which are encoded and mass stored in a device. [10][11][12][13] In addition, scores (similar to bonus, malus points) are given to the actual hypothetical retrosynthetic transformation step based on the reaction conditions and on acquired and stored strategic merits. As a very simple example, Figure 6 shows the symbolism that our mind recognizes as the reaction named "Aldol".…”

Intelligent predicting reaction performance in multi-dimensional chemical space using quantile regression forest

“…Machine learning (ML) can not only solve these problems well, but also find the relationship between data in a large amount of chemical information and help researchers to make reasonable judgments and decisions. ML has been increasingly favored by many chemists and has made progresses in computer synthetic design system [1,2], such as chemical reaction performance prediction [3][4][5][6][7][8], drug research and development [9,10], auxiliary high-performance materials design [11], auxiliary inverse synthesis analysis [12][13][14],…”

Computer Assisted Pathway Generation for Atrazine Degradation in Advanced Oxidation Processes

“…The fundamental methodology is to abstract chemical structure as graph and use graph theory to enumerate elementary steps of complex reactions. According to graph theory, each reactant and reaction mechanism can be modeled as a graph consisting of nodes (as atoms) and edges (as bonds) [5,6]. The addition and elimination of the edges could simulate the reactions.…”