New Neuroscience Initiative Could Help Build Precision Medicine Network

In the current medical practice environment, diagnosis and treatment decisions are based on patients’ medical history, coupled with an analysis of their symptoms.

In contrast, the promise of precision medicine could radically transform the existing model by collecting, integrating and analyzing comprehensive data across basic research and massive patient cohorts. All of this data will contribute to an interactive knowledge network, which will both inform basic research and guide precise diagnosis and treatment plans for each individual patient, empowering evidence-based decision making by clinicians and patients.

This topic was recently explored at the quarterly EVCP (Executive Vice Chancellor and Provost) Speaker Series, which presents internal and external leaders on key trends and ideas in science and medicine. Kate Rankin, PhD, an associate professor in the UCSF Department of Neurology, and Joe Hesse, director of technology and strategy in the UCSF Memory and Aging Center, spoke about the “Neuroscience Pilot of the UCSF Knowledge Network for Precision Medicine.”

UCSF is exploring the possibilities of precision medicine through a platform with six broad, overlapping elements. The University is investing institutional resources to advance each element of the platform and encouraging all faculty members to learn how they can contribute to and benefit from precision medicine. Bringing precision to the full spectrum of health is a key idea that emerged in the long-term planning initiative called UCSF 2.0, which calls for converting data to take new approaches to health, education and research.

One of the elements of UCSF’s precision medicine platform is developing a comprehensive data network that integrates research on the molecular mechanisms of disease with clinical data on individual patients. This UCSF Knowledge Network is being formed in two major areas: oncogenesis and cancer, and neurosciences and neurological disease, but information from all sources and disciplines will enlarge and enrich the network.

Rankin and Hesse are spearheading a neuroscience pilot initiative centered on linking datasets across levels and disciplines in a robust discovery platform that will represent a small-scale prototype of the precision medicine knowledge network, with the ultimate goal of building more comprehensive disease pathogenesis models to enable therapeutic progress.

There is already an enormous amount of neurological and other biological data that can be searched by researchers and clinicians. However, current computational tools typically allow only single-level visualization and analysis of datasets, with no possibility to connect different types of data in a comprehensive way and no user-friendly interface that could be used in real time by clinicians and patients to inform patient care.

In their pilot initiative, Rankin and Hesse propose to develop a platform for data integration and knowledge network development built on a linked knowledge base that could include multiple levels of neurobiological data, as well as neuroscience expertise. The data would range from genetic and molecular information through human neurologic, behavioral, and cognitive symptoms. Their effort does not aim at achieving a fully comprehensive knowledge network, but rather, testing the feasibility of various approaches and identifying potential roadblocks and solutions, to provide a model that can serve as foundation for a more comprehensive and clinically-relevant network.

The knowledge network will be populated with data from publicly accessible databases, as well as a rich dataset of multi-level human clinical information from local cohorts of patients with neurodegenerative conditions and healthy controls. The neurobiological knowledge network being developed by Rankin and Hesse is modeled on KBASE, a comprehensive knowledgebase for predictive plant and microbe biology designed by Adam Arkin, PhD, at Lawrence Berkeley National Laboratory.

In this knowledge network, clarifying organizational strategies, integrated analytics, and visualization tools will be applied to the data to drive towards dynamic, predictive models of function that can be used in both research and personalized clinical care. Despite its high level of sophistication and computing power, the knowledge network will remain a tool, while researchers will still be responsible for analytical decisions in the data mining process and clinicians will still be the final “integrators” of research knowledge and clinical care.

Rankin and Hesse emphasize that building the knowledge network and other elements of precision medicine will both stimulate and require deep changes in the culture, organization, operations and regulation of different disciplines, stakeholders, and professional bodies. In essence, the knowledge network aspires to bridge the current divide between clinical care and research, and to promote collaborations, while limiting “siloed” or – inaccessible – data pools and research across labs and disciplines in research and the clinic.

Moreover, it integrates the principles of open dissemination of scientific research, the acceptance of which will likely parallel the pace of changes in our approach to the scientific process. Thus, the knowledge network will not only revolutionize medicine by ultimately enabling everyday clinicians to access and apply the wealth of biomedical data into clinically actionable knowledge, but also form part of a large social experiment centered on data sharing and “open science.”