Twelve years ago, the Human Genome Project sequenced and mapped the human genome. The project enabled a massive leap forward in precision medicine — taking into account individual differences in people's genes, environments and lifestyles. Instead of going through the horrors of chemotherapy, which kills healthy cells along with cancerous cells, genomics can help patients by allowing for precision treatment plans tailored specifically to their disease.
[ Related: CIOs join forces to battle cancer ]
But so far, this type of precision medicine is available only to a select few, due partly to expense, but also to the amount of time involved. Today, it can take weeks or even months to sequence an individual's DNA, and then additional weeks or months to analyze it. Typically, about 1 TB of raw data comes off a sequencer, but it could be 2 TB or even 3 TB. That data has to be combined with imaging data and other clinical data.
"I experienced this first hand as a cancer patient five years ago," says Eric Dishman, director, Proactive Health Research, Intel. "The sequencing part only took a day in my case, but it took four months of processing and then another four months to use these huge files. I got lucky enough to survive that eight months from sequencing to treatment."
Intel and OHSU want to make it possible to sequence an individual's DNA, analyze it and formulate a precision treatment plan within 24 hours. At the Intel Developer Forum on Wednesday, the two partners announced an open platform-as-a-service (PaaS) solution called the Collaborative Cancer Cloud with the aim of getting to that point by 2020.
The Collaborative Cancer Cloud is precision medicine analytics platform that allows medical institutions to securely share patient genomic, imaging and clinical data. Dishman says it will enable large amounts of data from sites all around the world to be analyzed in a distributed way, while preserving the privacy and security of the patient data at each site.
"The challenge now is that as more and more cancer genomes and more genomes for all diseases get sequenced, we're facing a huge amount of data that's getting generated," says Dr. Brian Druker, director of the Knight Cancer Institute at OHSU. "We need to be able to share this data in a way that allows us to decipher what it's trying to tell us in a timely fashion.
"We need data shared securely in a way that accelerates our ability to do analytics and get treatments to patients," he adds. "We need a future where this becomes much more commonplace. Imagine where we'll be in 2020. Instead of two percent of patients getting sequenced, it will be 10 percent, 20 percent, maybe even 100 percent. All in one day."
Intel and the Knight Cancer Institute will launch the Collaborative Cancer Cloud and expect two new institutions to come on board by 2016.
"And from there, we can open up this federated, secure Collaborative Cancer Cloud network to dozens of other institutions — or let them create their own — to accelerate the science and the precision treatment options for clinicians and their patients," Dishman says. "They can also apply it to advance personalized research in other diseases that are known to have a genetic component, including Alzheimer's, diabetes, autism and more."
[ Related: Why Health Data Is a Big Data Challenge ]
Each partner will maintain control of its patients' data, Dishman explains. But the federation will allow the shareable cancer treatment knowledgebase to grow in availability. Essentially, rather than transporting massive volumes of data, the Collaborative Cancer Cloud will allow clinicians to run analytics on all of the data where it resides.
"We want to help harness the power of that data in a way that benefits clinicians, researchers and patients with a better knowledgebase and preserves security and privacy," Dishman says. "By securely sharing clinical research data amongst many institutions while maintaining patient privacy, the entire research community can benefit from insights revealed in large data cohorts."
As a complement, Intel plans to deliver open source code contributions to ensure the broadest developer base possible is working on delivering interoperable solutions. Dishman says open sourcing the code will drive both interoperability across different clouds and allow analytics across a broader set of data.
"That to me is the real power of academics working with Intel," Druker says. "The analytics clearly have to be improved and sped up. As cancer biologists, we know what we have to learn from the data. Intel computer scientists can help us optimize the code and get to more efficient analyses."