Names, chemical conversion

Automatic recognition and correction of these errors is a very important component of the chemical name conversion process. Based on available information, this procedure is implemented in the most flexible way in the CambridgeSoft Name=Struct program.48... 

Thus, the conversion of chemical names into structures consists of parsing the name into longest text fragments. These fragments are submitted to lexical analysis and the derived lexical units are compared with a collection of predefined units in a dictionary. 

Names like this are fine for familiar compounds that are widely used and referred to by chemists, biologists, doctors, nurses, perfumers alike. But there are over 16 million known organic compounds. They can t all have simple names, and no one would remember them if they did, For this reason, the IUPAC (International Union of Pure and Applied Chemistry) have developed systematic nomenclature, a set of rules that allows any compound to be given a unique name that can be deduced directly from its chemical structure. Conversely, a chemical structure can be deduced from its systematic name. 

The conversion of chemical names and identihers into appropriate chemical structure representations offers the ideal path for chemists and organizations to mine chemical information. Because chemical names are not unique and a multitude of labels can map to a single chemical entity, the facile conversion of alphanumeric text identihers to a connection table representation enables superior data capture, representation, indexing, and mining. The industry s need to mine more information from both the historical corpus as well as new sources is obvious, and a number of researchers have initiated research into the domain of chemical identiher text mining and conversion. Multiple efforts have been made in the held of bioinformatics research,8 and, while interesting as a parallel, in this chapter we will focus the efforts to extract and convert identihers related to chemical entities rather than, for example, genes, enzymes, or proteins. 

The scheme by which chemical structures are mined from chemical documents is shown in Figure 3.1. The greatest hurdle associated with successful mining of chemical structures via chemical N2S conversion is the quality and complexity of the chemical names themselves. Thus, a significant part of this chapter is devoted to the consideration of the quality of names and its contributions to the procedure of conversion. 

The first publication about the computer translation of chemical names was published by Garfield in 1961. In that article, he described the conversion of names into chemical formulas and initiated the path toward N2S algorithm development.45 Developments in 1967 at CAS provided internal procedures for the automatic conversion of CAS names into chemical diagrams.46 47 The first commercially available software program was CambridgeSoft s Name=Struct released in 1999,48 now patented,49 which was followed shortly by ACD/Labs ACD/Name to Structure product released in 2000.50 Two more commercial products are available Chemlnnovation s NameExpert51 and... 

The conversion of chemical names into chemical structures can be represented as two intersecting schemes utilizing a lookup dictionary and using syntax analysis. A combination of these two approaches is definitely needed for the analysis of chemical names in the real world. 

For the conversion of systematic names, a more powerful and flexible approach must be based on the parsing of the chemical names and the application of syntax analysis. Figure 3.4 illustrates the principle steps of this procedure. The first step in the process, lexical analysis, splits the whole chemical name into a series of name fragments, known as lexemes, that have structural or grammatical meaning. Also... 

Chemical nomenclature has a very large number of specific procedures to create chemical names, and many of these are not easily amenable to algorithmic representation, requiring significant investments in both development and validation time to develop automated procedures. Software developers of N2S engines prefer to support just the basic operations for conversion, at least at the early stages of development. 

This example demonstrates that one of the main challenges for an N2S conversion algorithm applied to data mining is the conversion of chemical names that are not strictly systematic, are ambiguous, or include typographical errors or misprints. 

One of the primary issues with systematic nomenclature is that some names can appear systematic in nature but, in fact, are not. They can have the expected structure of a chemical name generated according to a rules-based system but are false systematic names, at least in their specific context. When the N2S conversion algorithms... 

Table 3.3 shows that Name=Struct supports four main types of errors inside chemical names addition, deletion, replacement, and pair inversion. For the conversion of names generated by OCR, the most common error is character replacement. For example, the name heptane-2-car6oxylic acid shown in Figure 3.2 and resulting from OCR cannot be converted to a structure. 

Chapter 3 Automated Identification and Conversion of Chemical Names to... 

---