Health IT, Pharma

Cloud computing: A critical component of precision medicine

Phenotyping needs to catch up with genotyping: That’s one key move that must be made […]

Phenotyping needs to catch up with genotyping: That’s one key move that must be made for us to bring this precision medicine thing to widespread fruition. And to connect these dots most effectively, more robust cloud computing will be a critical component, some say.

President Barack Obama rang in a new era – that of personalized medicine – with his endorsement during his State of the Union Address this week.

“I want the country that eliminated polio and mapped the human genome to lead a new era of medicine—one that delivers the right treatment at the right time,” he said.

But what does this entail? Much like his 2013 BRAIN Initiative – which slated many millions and a focused federal effort on garnering a deeper understanding of disorders of the CNS – the details at the outset of this effort toward targeted therapy are, well, squiffy. However, this we know:

The cost of sequencing’s lessened considerably, but we still don’t know exactly how to make sense of this information – it’s just too much data to trawl through at the moment.

“The issue isn’t just collecting human genomes, but having rich enough data about patient behavior and the phenotypic data from EMRs, so we can integrate health data to form a rich picture about a patient,” said David Shaywitz, the Chief Medical Officer of Bay Area genomics cloud computing startup DNANexus.

This was actually a prevailing theme at last week’s JP Morgan conference.

“The sequence part is the tool, the commodity – but it’s everything else you’ll do with it that’s the secret sauce,” Regeneron chief scientific officer George Yancopoulos said at the time.

Nature recently published a piece on the importance of shifting from sequencing to analyzing function. The Cancer Genome Atlas project, a $100 million effort launched in 2006 to genetically profile 10,000 tumors, just wrapped up – but scientists are now realizing now they’ve got a glut of data, more emphasis needs to be placed on interpretation. And, of course, we’ve got to find ways to consolidate the reams of big data being spun out on any given individual:

The NCI is also backing the creation of a repository for data not only from its own projects, but also from international efforts. This is intended to bring data access and analysis tools to a wider swathe of researchers, says Staudt. At present, the cancer genomics data constitute about 20 petabytes (1015 bytes), and are so large and unwieldy that only institutions with significant computing power can access them. Even then, it can take four months just to download them.

Stimulus funding cannot be counted on to fuel these plans, acknowledges Staudt. But cheaper sequencing and the ability to use biobanked biopsies should bring down the cost, he says. “Genomics is at the centre of much of what we do in cancer research,” he says. “Now we can ask questions in a more directed way.”

“The rest of the world understands the potential of the cloud – the NIH is just contorted over coming up with some sort of policy about it,” Shaywitz said.

Storied San Diego geneticist J. Craig Venter also buys into this idea of harnessing cloud computing to make sense of sequencing data. He said this as part of a July Congressional testimony:

There are two thresholds we just passed that actually allowed us to form human longevity. One was the sequencing technology that just barely passed the threshold of cost and accuracy, but the most important changes are in the computer world. And we are going to rely very heavily on cloud computing, not only to house this massive database, but to be able to use it internationally. We will have operations in different parts of the U.S. and even in Singapore to allow us to do computes 24 hours a day. The cloud sort of makes that seamless instead of trying to transport this massive amount of data, trying to move things from my institute in Rockville, Maryland, to La Jolla, we had dedicated fiber but it is
now so slow with these massive data sets, we use Sneakernet or FedEx to send disks because we can’t send it by what you would think would be normal transmissions. So the use of the cloud is the entire future of this field.

 

Shares0
Shares0