At Oracle, it’s Susie Stephens’ job to figure out how to make databases and application suites more relevant to customers in life sciences. As you can imagine, Stephens is a very busy woman.
Stephens, principal product manager for life sciences at Oracle, and her group have spent the last three years making the database company a known name among life science labs. The primary challenge areas they identified were (and still are): accessing distributed data, integrating various types of data, handling a large volume of data, enabling secure collaborations, and finding patterns and insights, Stephens says.
A major triumph for her team came in January of this year with Oracle’s release of 10g, a database with specific applications and capabilities — such as Blast, support for expression searches, and non-negative matrix factorization — targeted at scientists and informaticists in this market. Now, Stephens is busy getting the second version of the database, known as 10g R2, into beta testing, while at the same time working on the second generation, known around the company as 11g.
At press time, few details were available for R2. Stephens says new or changed functions are confidential until after the beta phase, and that there’s no predefined time line for that. 10g was in beta for “over six months,” she says, noting that that was “a longer beta test process” than usual “because we wanted to make sure 10g was a very solid release.” Though she couldn’t talk about specific functions, Stephens did say that Oracle customers have expressed interest in various analytics and new algorithms they’d like to see in the database.
So far, she says, user group meetings this year in London and Reston, Va., have been fairly positive. The most popular tutorial sessions have been “managing images in the database [and] using Blast functionality,” Stephens says. Blast, for instance, can be run on 10g with a multilayered query search. Other analytics, meanwhile, enable work like building data mining methods that would go through unstructured data from a clinician’s report as well as gene expression data, she adds.
During the years she’s been studying this field, one change Stephens has seen has been a shift away from its former high-performance computing and supercomputing focus. “We mainly have people talking to us about integration” — of data, applications, or both — “rather than high-performance computing,” she says. “We see that more people are moving to the compute farm or grid environment where they can use commodity hardware.”
— Meredith Salisbury