By Allison Proffitt
February 28, 2012 | Thomson Reuters announced a series of APIs (application programming interfaces) for its Cortellis life sciences platform this morning called Cortellis for Informatics. The APIs will give customers direct access to the Cortellis content.
The product grew out of an exchange with Thomson Reuters customer Biogen Idec. William Hayes, director of decision support at Biogen, explains that data wasn’t the problem. “We have plenty of information,” he says. “More information than we could possibly use!” Biogen Idec initially approached Thomson Reuters asking about ways to cull specific data from the Thomson Reuters databases and deliver them to teams internally.
“We wanted [to pull] selected segments of the Thomson data into our system for customized competitive intelligence,” explains Hayes. Biogen’s initial plan was to Web scrape the Thomson data—“With permission!” he says—but it was tedious and not very effective. “It was very problematic, trying to pull 200 different compounds, it would take either a long time or fail and required lots of care and maintenance.”
An API was the natural solution. “From the beginning we said [APIs] would be a lot easier,” says Hayes. “But it takes time to build a decent API… An API on top of content allows us to customize what we’re pulling out and deliver internally [by email or RSS] bits of information someone can quickly review and decide if they want to follow up.”
The idea aligned with Thomson Reuter’s aims as well. One of Thomson Reuter’s goals is to enable analytics against content rather than just publishing it, says Wendy Hamilton, senior vice president at Thomson Reuters.
Thomson Reuters Cortellis is a brand new platform that integrates data that previously existed in several legacy databases that didn’t talk to each other—including drug monographs, company records, patent reports, deal reports, and more—she says. The Cortellis platform includes a Web portal, but the Cortellis for Informatics APIs are a “significant step” toward “pushing content into user workflows,” says Hamilton. She echoes Hayes’ concerns. Researchers are busy; they don’t always want a lot of Web portals. They need information at their point of need, she says.
The Cortellis for Informatics APIs have been developed around areas of content in the Cortellis platform including investigational drugs, targets, patents, analytics, and ontologies. An API on clinical trials is expected by the end of March, and more will be released by the end of the year. APIs will be offered as more data and databases are integrated into Cortellis. Hamilton mentioned offerings for PK data, omics data, pharmacology data and molecular data as likely areas of expansion.
Cortellis access is a subscription-based service that is hosted by Thomson Reuters and that enables people and companies to access rich content either through a Web interface or programmatically via Cortellis for Informatics.
If a client chooses the API route, an API “key” gives the client’s developer team access to both SOAP and RESTful web services, allowing them to query Cortellis content programmatically. Developer sites provide documentation as well as “tips and tricks” to get people started. Clients also have the option of choosing from a selection of add-ons readymade for the popular components including Microsoft SharePoint and Accelrys Pipeline Pilot, or to make use of Thomson Reuters’ professional implementation service.
It’s, “a very comprehensive, excellent product,” says Hayes. “But it requires a lot of expertise to use. The more we can extract and simplify, the wider the audience can use the information.” Since their first foray into Web scraping, Hayes and Biogen have converted to the APIs and, “It works much better!” It’s being used by both research scientists and marketing and sales staff.
“The more we can deliver what our end users within the company need, the more attached we are to their system,” Hayes says. “Divorce gets harder and harder.”