NEW YORK (GenomeWeb) –RowAnalytics, the Martinos Center for Biomedical Imaging, and Envision Genomics are partnering on a new initiative called Giro Health, which they describe as an open initiative in precision medicine that seeks to improve diagnosis and personalize treatments for complex neurological and neurodegenerative diseases.
The terms of the agreement allow the partners to use their respective technologies and expertise to analyze clinical genomic data and multi-modal imaging data to support efforts to develop new treatments for conditions such as multiple sclerosis, dementia, Parkinson's disease, and epilepsy. The partners also seek to help clinicians better diagnose instances of these diseases more accurately as well as select more appropriate therapies and trials for their patients. They also plan to provide personalized tools for patients that will help them avoid adverse events and change aspects of their lifestyles to improve their overall health.
"Neurodegenerative diseases are some of the most complex; there are many different causes that have to be deconvoluted," RowAnalytics CEO Steve Gardner told GenomeWeb in an interview. Unpacking the roots of these conditions will involve integrating and analyzing large quantities of both genomics and multi-modal imaging data and require informatics capabilities that all three partners bring to the table.
In addition to genomic and imaging data, the partners also plan to pull in clinical and epigenetic data as well as some lifestyle data into their analyses. In some cases, they'll also have access to environmental data as well as sensor data that will provide information on patient behavior.
"We want to provide new R&D directions for those [disease] areas that are not very well served [such as] ALS and Alzheimer's," Gardner said. The benefits to patients include providing clinicians with the tools they need to evaluate an individual's personal risk of developing neurodegenerative disease.
The partners will also develop mechanisms to provide information to patients via their mobile phones, for example, which could help them improve their lifestyle and make modifications to their behaviors that could alleviate their symptoms. "That's partly about making sure that the combination of drugs that they are on is right for them and it's partly something as simple as making the right food choices to make sure that the ingredients in the food do not exacerbate their disease," or cause adverse drug-food interactions, Gardner said.
For its part, the Martinos Center for Biomedical Imaging brings significant expertise in high-resolution multi-modal imaging techniques including functional magnetic resonance imaging technology that makes it possible to generate detailed images of the brain. Meanwhile, Envision Genomics provides whole-genome and whole-exome sequencing services to clinicians and their patients.
For its part, RowAnalytics is providing its precisionlife platform, which provides tools and capabilities for rapid and parallel analysis of various forms of structured and unstructured data including medical images, videos, clinical reports, and whole genome. In addition to spearheading the newly launched initiative, RowAnalytics is preparing to formally launch various portions of its flagship product starting early next year.
Last week, the company opened its first US-based office in Cambridge, Massachusetts, from which it will run its commercial and technical operations in the US. This will include providing onsite assistance to customers who are using its platform for biomarker discovery and genome-wide association, drug repurposing, and patient stratification studies. The company has already begun adding to its headcount in the US by hiring a new chief operating officer and a vice president of market development and it will continue to hire additional staff in the coming months.
The list of tools includes a knowledge management platform called precisionlife Search that supports full semantic indexing and supports easy maintenance of ontologies and knowledge graphs. Users can search by keywords or they can cut and paste whole sentences or paragraphs or match whole articles. Over time, the system learns users' personal interests and is able to send alerts automatically when new, relevant papers are published.
Other products available on the platform include precisionlife Markers, a biomarker discovery solution that uses a massively parallel algorithm to link disease risk to specific features in genomic, phenotypic, clinical, and imaging datasets. The software enables the analysis of biomarker networks with up to 50 genomic, phenotypic, and clinical factors acting in combination; provides semantic annotation for validating and interpreting new biomarker networks; and supports patient stratification for clinical trials design and healthcare analytics, according to the company.
Another tool, precisionlife Diet, offers personalized dietary advice via a secure application programming interface or an app on the patient's phone. This includes providing fully personalized advice based on over 500,000 drug-drug, drug-disease, and drug-food interactions updated regularly as new nutrition and pharmacology data sources are published.
"We are trying to look at not just different data types information … but to look at it en masse from large populations and then distill it down so that it's a meaningful treatment plan or meaningful information to an individual," Loralyn Mears, Row Analytics' vice president of market development, said in an interview. "It's going from the population to the personal N =1 level."
This breadth of data that RowAnalytics' platform can take in was a key factor in IBM's decision to work with the company to pair its software with IBM hardware, according to Janis Landry-Lane, a global sales executive for IBM's high-performance computing business. The two companies have a business relationship that allows them to co-sell RowAnalytics' platform with IBM hardware. Landry-Lane also played a key role in bringing together the partners of the GIRO Health initiative.
"They had a comprehensive approach to disease," she said. "They addressed the scale of analyzing large-population, multi-modal data and delivering the results for a comprehensive individual treatment plan. We found their approach unique, [and] we wanted to address the computing requirements for the complexity of their algorithms with the massive amount of data being analyzed."
According to the results of internal tests that ran the precisionlife Markers software on data from a genome-wide association study, the researchers were able to analyze combinations of two, six, and 17 SNPs in 12 minutes, six hours, and six days, respectively. The company ran the software on an iteration of IBM's reference high-performance computing architecture for healthcare and life sciences. The particular iteration used here also included GPUs that helped boost performance, according to Landry-Lane.
"Our role is to create an architecture which really … maximizes the utilization of the existing compute hardware as well as an architecture where this platform can scale seamlessly," she said. The system also includes an archive solution for longer-term data storage that allows easy access to older datasets at any time. "Users see their data as if it is sitting on the disk, but it actually may be in an archive, [so] they can always get access to their data."
In terms of the GIRO initiative, Landry-Lane said that IBM's reference architecture for high-performance "big data" analytics in healthcare and life sciences can work with the applications that the partners provide. The reference architecture for GIRO Health will be built on IBM's Power Systems, which includes GPU acceleration and the scaling capabilities.
In deciding which HPC components to include, "we had to build something that can address many types of programming models and applications. We utilized some of the best practices that we have optimized for our largest high-performance computing customers," she said. "As we look ahead, we hope to have the privilege of supporting GIRO's combined efforts with their projects."
RowAnalytics and its partners in the GIRO initiative are seeking additional collaborators including research foundations, disease specialists, and clinical researchers.
"We are expecting to work with a number of different disease charity-sponsored projects [as well as on] some company projects," Gardner said. "There are a number of [groups] who have collected large datasets but lack the ability to understand the details of them, [or in other cases] they've got some of the data, maybe they now need to bring in imaging."
So far, the partners have a shortlist of five or six projects that they are mulling over for their first steps.
The partners have secured some funding, which they will use to start work on some projects, and have already secured some datasets and are working on a number of proofs-of-concept. But they are seeking additional funding partners as well as research collaborations. In fact, "one of the reasons for doing the launch is to enable us to go out and talk to other funding partners," he said.
GIRO Health will be operational in the fourth quarter of 2017 and will be based initially at the Charlestown Navy Yard in Boston.