Nippon Shokubai is working to develop data science talent to speed up research and development, and has adopted JMP, which is easy to use even for statistical novices. However, using JMP to handle molecular structures is difficult and requires programming, which is a hurdle for beginners. Therefore, in order to learn the know-how of handling molecular structures in JMP, we held a Materials Informatics (MI) study session in cooperation with Mr. Katsumura of SAS Institute Japan Co., Ltd. At the study session, we shared an add-in (Toolkit for Materials Informatics) for handling compounds in JMP and analysis examples.
The add-in developed by Katsumura makes it possible to visualize molecular structures and extract features without programming. This function will be introduced with a demo. In addition, an analysis example of "searching for chemical structures that satisfy desired property values" will be introduced. A model was built using the add-in and standard JMP functions, and commercially available compounds were virtually screened.
These efforts are expected to further speed up research and development by making it possible to use JMP to analyze compounds that previously required programming.
