We have been doing studies on the use of information technology in chemical data analysis and drug development study. When you think about the use of higher computer technology in chemistry and the related field, you should notice that there are domain-specific problems that must be solved at the same time. For example, one of the most important problems is chemical structure processing by a computer. To aim at the realization of a chemical artificial intelligence, we have to establish dedicated information technology for the chemical structure processing and the interpretation of a chemical structure. This article introduces some examples of the utilization of artificial intelligence in drug development on the basis of machine learning. They include applications in both supervised learning and unsupervised learning.

The therapeutic and adverse effects of pharmaceutical medicines are subject to marked individual variations. Variations in drug metabolizing enzymes could result from genetic polymorphisms such as nucleotide insertions, deletions, and substitutions, and copy number variations. Understanding the molecular mechanisms underlying these genetic differences would help to develop personalized drug therapies where medications are optimized and doses are selected according to the patient's genetic information. We have comprehensively identified the molecular properties of around 300 genetic variants of drug-metabolizing enzymes. Using these results together with the recently discovered 2,049 Whole-genome Japanese Reference Panel and Bioinformatics, we analyzed the enzymes' 3-dimensional docking simulations in an attempt to understand several rare variants of drug-metabolizing enzymes that have been previously overlooked owing to their low frequency. The findings we obtained will be extremely useful in analyzing “real-world data” in clinical studies.

Open Science is a concept that includes open research data and open access to scholarly literature, and it has been promoted vigorously by government organizations and scientific communities. Utilization of published research data by not only researchers but also corporations and citizens is expected to lead to economic effects resulting from employment and innovation and to solve social problems. This paper outlines policy trends related to open science, provides examples of utilization of open research data, and elaborates the direction of scholarly communication. Further, it introduces barriers to data disclosure, academic libraries’ supports to researchers, and efforts to solve problems by the community.

The Unified Medical Language System (UMLS) is available from the National Library of Medicine (NLM) site. It's called Meta-thesaurus. Over 200 terms are included in the UMLS. All terms are associated with a concept. We focused on this point. They are not only associated with other terms but also with a database. We found the potential for one term to open the door to the next.