Donate to Science & Enterprise

S&E on Mastodon

S&E on LinkedIn

S&E on Flipboard

Please share Science & Enterprise

Harnessing Big Data for Precision Medicine

Big data graphic

(DARPA, Wikimedia Commons)

13 February 2016. Precision medicine aims to match individualized genomic data with therapies to provide personalized treatments for people with disease. A panel today (13 February) at the American Association for Advancement of Science or AAAS 2016 annual meeting, described how big data — finding insights in large data sets — help make precision medicine possible, with Lyme disease as a case study.

Lyme disease is transmitted to humans by bites from the black-legged or deer tick that carries Borrelia burgdorferi bacteria, with symptoms including fever, fatigue, headache, and rash. The disease can be treated with antibiotics, but if left untreated, infections can spread to the joints, heart, and nervous system, and become chronic and debilitating. Centers for Disease Control and Prevention says between 19,000 and 30,000 cases of Lyme disease are reported in the U.S. each year, mainly in northeastern states and the upper Midwest, although new data reported at the AAAS meeting suggest that number could be as high as 300,000 per year.

Because Lyme disease is a complex disorder, often difficult to treat and diagnose in its chronic state, said the panel, it is a good candidate for precision medicine. Lorraine Johnson, a Lyme disease survivor and CEO of in Los Angeles, pointed out that traditional research models are not working for Lyme disease, and registries of data from patients can help make the disorder a more visible target for academic and industry scientists.

Johnson called Lyme disease, a “research disadvantaged disease,” meaning it’s generally ignored by pharmaceutical companies, and it is too prevalent a disorder to be classified a rare or orphan disease. While there may be more people with Lyme disease than HIV/AIDS, said Johnson, only three clinical trials of new treatments were undertaken so far, none of which were funded by industry. Conventional clinical trials have strict screening requirements, which limits participation to a relatively few patients, suggesting that those taking part in the trials may not represent the general patient population.

In addition, Johnson noted that conventional trials are “not messy enough” to address the complexities of Lyme disease. Clinical trials generally test one intervention for relatively short periods of time. Lyme disease populations are highly diverse, with those diagnosed early often much different from those whose disease is diagnosed late, and treatment effects often varying widely.

Instead of a few clinical studies with a few participants, Johnson called for large numbers people with Lyme disease to band together and offer their data in a way that gets the attention of the research community. Her organization started MyLymeData, surveys of people with Lyme disease to pool their diagnosis and treatment experiences. MyLymeData, said Johnson, is “patient-powered research” that can reveal disease patterns, identify subgroups, explain differences resulting in slow and non-responders to treatments. She said MyLymeData now has 3,000 participants, and her group is aiming to sign up a total of 10,000.

A SLICE of big data

John Aucott, Director of of the Lyme Disease Clinical Research Center at Johns Hopkins University School of Medicine in Baltimore, showed how big data can enable precision medicine for treating Lyme disease. Aucott gave results from the Study of Lyme Disease Immunology and Clinical Events or SLICE to understand why some people with Lyme disease develop a post-treatment syndrome that can last years, while others respond quickly to treatment and avoid the long-term condition.

The study followed 29 individuals with the disorder for 2 years, amassing large sets of clinical and biological data. To this collection, Aucott and colleagues added DNA sequencing, gene expression, blood protein, and cell analysis from patient blood samples. The results enabled the researchers to identify differences in immune  responses of participants, which suggested different types of treatment strategies.

Beyond the immediate results of the study, Aucott pointed out that big data approaches to understanding Lyme disease and other disorders are readily available, in some cases to anyone with a Web browser. For example, said Aucott, Google Trends reports of Web searches for Lyme disease and tick bites show seasonal patterns similar to epidemiology data. In addition, databases of insurance claims reveal one-third of patients with Lyme disease develop post-treatment symptoms, such as fatigue.

Aucott added, however, that a key requirement for big data investigations is interoperability of data. He plans to extend his analysis to other hospitals in the Johns Hopkins University network, but to really understand Lyme disease, the investigation has to be nationwide. Aucott would also like to extend the research model to other complex diseases, such as chronic fatigue and fibromyalgia, but interoperable data will be required.

D.J. Patil, the U.S. government’s chief data scientist, described the role played by big data in the government’s precision medicine initiative, announced a year ago by President Obama. Big data, Patil noted, enables individuals to contribute their genomic data and electronic health records, thus taking part more actively in their own treatments.

The government’s initiative aims to collect data from the analysis of 1 million individual human genomes from sources such as the Veterans Administration, as well as private sources like Having these data encourages explorations that reveal correlations and insights not previously evident. Patil underscored the importance of data ethics from the start, to make the data open and available for research, while at the same time protecting the privacy of individuals.

Read more:

*     *     *

Comments are closed.