“When we saw a patient, we were often scrambling,” says Kate Rankin, PhD, a clinical scientist and neuropsychologist at UCSF’s Memory and Aging Center. She’s describing the process she’s used, until recently, to sort out the relevance of a patient’s clinical symptom scores using calculators and reference charts. “We were literally doing the math by hand, on the fly, during the patient’s visit.”

But in fall of 2015, Rankin’s ritual of scrambling and calculating is being significantly upgraded. She and Joe Hesse, the center’s Director of Innovation and Strategy, are launching the dashboard functions of their Knowledge Network Core Technology system. KNECT, as it’s known, is a data management and computational platform that can harmonize a slew of different types of data, serving up an analysis that’s as meaningful to a researcher as it is to a physician sitting with a patient. As such, it is a prototype of precision medicine in action.

“This approach simultaneously improves our clinical decisions and makes research easier,” she says. “We’re bringing some of our best science right to the bedside.”

A clinical researcher, Rankin spends her days thinking about and imaging the brains of aging patients with neurodegenerative disease. Her specialty is visualizations that combine MRI brain scans in a way that allows her to compare neurologic damage in different patients and identify patterns of disease.

Rankin sees the ideal knowledge network not just as a dynamic means of sharing these brain images – or any other data – but also as a tool for contextualizing that data so it has meaning to researchers and practitioners across disciplines. In her office, she pulls up a screen with information on a 65-year-old male patient on whom she’s done cognitive testing, identified simply by a number. The screen shows the patient’s cognitive data and gives ranges for values considered normal and values that may be cause for concern. She can also see exactly what parts of his brain are different from other 65-year-olds. Any clinician or researcher with credentialed access to KNECT can incorporate this new information into their assessment of the patient, whether or not they are experts in brain imaging or psychometrics.

As Rankin puts it, “We’re building user interfaces that let people run analyses they haven’t been able to run before and see data they haven’t been able to see before.”

That can also accelerate research. She’s often aware of brain scan and behavioral data for her research patients, but doesn’t have access to their genetic data, or lab data other researchers have collected. “Maybe patients with this genotype just went through a sleep study. I’d love to look at my social and emotional behavior data against their sleep and inflammation data.” KNECT can tell her whether that data exists in the system and integrate it with hers. And as more researchers and clinicians add data to KNECT, these opportunities will expand.

Typically, Rankin points out, sharing datasets relies on colleagues running into each other in the hallway or at a conference. It also often involves having to bring in a programmer to find an elegant and accurate way to mesh data from different sources. KNECT provides the technological anchor for that exchange and is tackling many of the problems of harmonizing data sets more systematically.

But Rankin and Hesse know that not every researcher is going to leap on the data-exchange bandwagon. Science is a competitive landscape, in which data is a valuable resource. To allay researchers’ concerns, the team is building in an intricate auditing system. “There is very careful management of provenance: Where did this data come from? Who has looked at it? What have people done with it? Who has downloaded it?”

The ultimate vision for KNECT goes far beyond the scope of what Rankin and her programming team have accomplished at the Memory and Aging Center. So far, the system is a proof-of-principle – more a prototype than an end product. Ultimately, they’d like to find a company to collaborate with UCSF and grow the prototype into a large-scale framework. Ideally, people throughout UCSF and beyond would be able to work, share and collaborate within this environment.

To fully realize precision medicine, the knowledge network will become even broader, encompassing medical and research centers throughout the world. Rankin, who’s been thinking about ways to solve this puzzle for a decade, acknowledges that it’s a monumental task. But, she argues, that’s not a reason to stop moving forward.

“The National Academy of Sciences report that outlined the knowledge network we need for precision medicine likened it to building the great cathedrals of Europe,” she notes. “They were started by one generation and finished by another.”

Like those cathedrals, the knowledge network is being constructed brick by brick. Rankin has no doubt that there will be more generations of KNECT and more generations of researchers building on it.