Screening Methods for Drugs Using Chemoinformatics Methods for Beginners

Since the term informatics first appeared in the world of chemistry in the late 1990s, many new methods have been developed, as described in this book. However, chemoinformatics has not yet been widely used in practical screening studies. The reasons for this include the following. Compounds are diverse and big data do not exist. Chemical data can vary widely in value depending on the data source. Development goals are sometimes vague, such as “moderate solubility”. With regard to toxicity, its situation is complicated by species-dependence, but even animal testing is becoming unacceptable. In addition, many of the chemoinformatics tools discussed in the other chapters of this book can be somewhat too advanced for beginners to start with. While it is appropriate for beginners to start with multivariate analysis and basic machine learning, there are few examples of systematic study of compound design. In this chapter, the author shows step-by-step how cheminformatics methods can be used in the screening of candidate compounds for inhaled anesthetics. The first step is to estimate physical properties from molecular structures. These physical properties are then used to screen drug candidate compounds. Methods for semi-automated reverse design of candidate compounds will also be described.