Thanks for visiting Research Information.

You're trying to access an editorial feature that is only available to logged in, registered users of Research Information. Registering is completely free, so why not sign up with us?

By registering, as well as being able to browse all content on the site without further interruption, you'll also have the option to receive our magazine (multiple times a year) and our email newsletters.

Turning names into structures simplifies drug discovery

Share this on social media:

Topic tags: 

A tool to normalise information about chemical compounds is the first result of a partnership between Elsevier MDL and Temis

Knowing what has been made and tested before is an important part of the drug discovery and development process. With so much information online, it sounds like a simple matter to search and find what has been published about particular substances in the past. The reality, however, is more complicated. People describe and refer to chemical compounds in a number of different ways and a simple search tool has no way of knowing that a particular brand name or commonly used name corresponds to the standard chemical nomenclature.

To tackle this problem, Elsevier MDL, which provides chemical search tools, has teamed up with Temis to turn the disparate chemical names into easy-to-interpret chemical structures.

The French text-mining company has been a supplier to US-based MDL since 2000. MDL used Temis's Insight Discoverer Extractor engine in its Chemical Name Recognition software for extracting chemical names and reaction schemes from patent databases and scientific literature. MDL's Patent Chemistry Database backfile, which uses this software, was launched two years ago.

After this launch, the two companies realised that the technology used to extract the information from patents would be useful in other settings and they decided to release it as a tool that people could use with their own data. As Guillaume Mazières, Temis's vice-president of sales and marketing, explained: 'We switched from a contractor/supplier relationship to a real partnership'.

The result of the partnership is the Chemical Entity Relationships Skill Cartridge, which works with Temis' text mining engine to extract chemical information. The target customer base is the life science industry, as this is already a strong market for both companies. Typical users fall into two groups, according to Mathieu Plantefol, Temis's life science consultant. The first consists of people from intellectual property departments in life science companies. They would use the tool to make sense of patents by pinpointing relevant names and formulae to establish whether something can be patented and what has been done already. The Skill Cartridge normalises the range of data available by converting names of compounds into chemical structures.

The second group of users contains those who evaluate new types of compounds that might be biologically active. The new tool can be used in conjunction with other Temis products such as the Biological Entity Relationships Skill Cartridge and the Medical Entity Relationships Skill Cartridge. These tools use the text mining engine to, for example, identify diseases or proteins.

The new Skill Cartridge is already attracting interest in the life-science industry, according to Temis. It is currently being evaluated by three major customers and the first customers in the US and Europe are expected by the end of the first quarter of this year. 'The response has been very strong; the tool does not really have any competitors,' said Mazières. He believes that the partnership between Temis and MDL is a strong selling point with customers and says that the two companies plan to work together in the future. And this could extend beyond Temis's established life-science market. 'We have had some requests to open it to other types of companies,' commented Mazières.

The Chemical Entity Relationships Skill Cartridge

The Chemical Entity Relationships Skill Cartridge from Elsevier MDL and Temis is a new software application that identifies chemical compound names, chemical classes and molecular formulae in text documents and translates extracted information into chemical structure.

Its features include classification and relevance weighting, where identified chemical terms are assigned to specific chemical concepts according to semantic categories such as regular chemical name, chemical class name and trivial name.

A name-to-structure translation service enables the user to automatically match textual information with proprietary chemical structure libraries.

A unique fingerprint is provided for each identified structure for de-duplication purposes.

Contact: Martine Falhon; martine.falhon@temis-group.com

Siân Harris