Vol.4 No.1 2011

47/78

Research paper : Development and release of a spectral database for organic compounds (T. Saito et al.)−44−Synthesiology - English edition Vol.4 No.1 (2011) imitations and of others making profit by using the imitated spectral information of SDBS. The digital data have high workability so reproduced materials from the data have a high commercial value. If most of the digitally compiled spectral data were copied by others, they can construct a spectral database similar to SDBS or may be able to create more valuable database from the data. This is a menace for SDBS. We think even if the copyright infringement is recognized and can be appealed in court, we must spend considerable time and effort on the suit. Large amount of data copy is a copyright infringement even if the data is in GIF format and may cause SDBS to be in a threatening situation. Therefore, user access from the Web is monitored all the time. When a third party wants to use data in the digital or the image formats, the licensing from AIST is necessary. In the case where the party wants to sell the data, they have to pay a royalty to AIST. We think this is a different point from the copyright in disclosing the data in the Web. We are not allowed to name names because of secrecy agreements, but we have several experiences of offering a large amount of data at a time, and in each case we have received payments of royalties based on the contracts. In the case of IR spectra, we have a contract to offer data with royalty when we release new spectral data. We also have experience of offering GIF image data to a database operated in the United States of America. 6 Comparison with other spectral databasesQuestion (Akira Ono)I think there are spectral databases besides this one in the world, especially, those distributing the digital spectral information to private companies with a fee operated by private companies. Would you introduce such databases, and instruct us of the differences in roles and characteristics from this database?Answer (Takeshi Saito)There are not many spectral databases that are open through the Web. When looking at such free databases opened to the public, the number is very limited. We have not encountered a free accessible spectral database with this many 1H NMR spectral entries with their spectral patterns and chemical shift assignments.One of the free accessible databases on the Web is NIST Chemistry WebBook (http://webbook.nist.gov/) operated by the National Institute of Standards and Technology, NIST. This database compiles a variety of physical chemistry data as well as spectral data. This database is similar to RIO-DB of AIST because most of the data compiled in the WebBook are based on the research products of NIST. Besides 15,000 MS and 16,000 IR spectra, the WebBook compiles other spectra like an ultraviolet-visible absorption spectrum and a terahertz spectrum. There is no need to install software to browse the data of the WebBook. Spectra and other information are compiled along with the compound information, which you can search. We do not have a list of compounds compiled in the WebBook, however, it is our presumption that many generally used chemical reagents are compiled. This database works much like our database that serves as that of a public organization. MS data was sold as NIST 08 Mass Spectral Library which was released in 2008. Many spectral data were obtained at National Institutes of Health (NIH) and United States Environmental Protection Agency (EPA) and were evaluated at NIST. The data can be purchased for use on an independent personal computer (PC). The number of data compiled in this database is about 220,000 spectra for 190,000 compounds, which is much more than those open through the Web. When we did collaborative research with NIST during our former Agency of Industrial Science and Technology era, many of our MS data were compiled in the NIST MS Library. These data must have been included in the NIST Mass Spectral Library. This data library can be searched with a fee with MS spectral patterns obtained by many kinds of MS spectral instruments.SpecInfo releases data through the Web with access charge, and it consists of 90,000 1H NMR, 300,000 13C NMR, and many multinuclear NMR, IR and MS spectra. Data update does not seem to have been performed since 2006.Looking at a domestic site, MassBank (http://www.massbank.jp/) which is a high resolution mass spectral database of metabolites is open free to the public. As of April 05, 2011, almost 31,000 spectra from 20 research institutes have been compiled. Target compounds are specified and limited to metabolites. Users of this database should download special software for this database, and install it on their computers. This software enables you to search, browse, display full and expanded spectrum, and register spectral data to the database. The database construction was initiated and has been supported by the project, “Integrated Database System for Metabolomics”, under the Institute for Bioinformatics Research and Development of Japan Science and Technology Agency since 2006. As our database started with a research project and had developed into a solid base of our activity, we look forward to seeing how MassBank will be developed after this project is finished.Databases from Bio-Rad are using a non web platform. The database covers data of SpecInfo, NIST MS Spectral Library in addition to Sadtler spectral data. Our NMR spectral data are also compiled. This database is sold as a package of data and software for handling the data, “KnowItAll”, that needs to be installed on a computer. The compiled data are about 50,000 1H NMR, 430,000 13C NMR, 190,000 MS, 7000 Raman, and 230,000 IR spectra. Of our data, 13,000 and 11,000 1H and 13C NMR spectra, respectively, are included. Users can use a spectral pattern for spectral search, search for mixtures, and more complicated searches can be performed with the software. Sigma-Aldrich sells “Sigma-Aldrich Spectral Library” which has a total of more than 50,000 compounds of NMR, IR and Raman spectra. There are two formats of data on a computer and as a book.When comparing a database operating on the Web and on an independent PC, there exist both advantages and disadvantages for each. For example, an advantage of the Web based database is instantaneousness. Our database can perform data addition and update easily; our database is updated twice every year to supply new data to our users. Many of the other databases introduced here do not seem to update their data that often. On the other hand, a database on the PC has better usability. For example, spectral pattern matching search can be done which many users find useful.

※このページを正しく表示するにはFlashPlayer9以上が必要です