Forget Smartphones—Genomic Data is the Next Big Deal, Says Edico Genome CEO
February 3, 2014
By Josh Baxt, BioSpace.com News Staff
Genomics is difficult. First there are the biological hurdles, how to actually sequence a person’s entire genome. Then there’s dealing with all that information. How do we interpret terabytes of genomic data?
Universities and large corporations can crunch the data, albeit expensively, with server farms and supercomputers, but what about a hospital or lab? Owning or renting servers is a costly proposition and not every facility can afford to hire their own bioinformatics wizard.
“We’ve created a more specific platform, with new algorithms and a dedicated processor that’s designed for one particular job,” says CEO Pieter van Rooyen. “That’s one of the keys to take sequencing to the next level: making the whole workflow a lot simpler.”
According to Edico, a whole genome sequence can be analyzed with Dragen in around 20 minutes, as opposed to 24 hours with a conventional server. In addition, the savings on IT infrastructure could be as much as $6 million over four years.
A Tough Market
At this point, Edico is marketing a unique genomics-focused processor. Other companies, such as DNAnexus, Genalice and GenomeCloud offer software or cloud-based solutions. On one level, they are competitors, but van Rooyen believes they could also be customers, as Dragen could potentially accelerate their processes.
However, market acceptance may prove a barrier for Edico, notes Michael McManus, senior vice president at Knome, which provides hardware and software to help interpret genomic data.
“Edico has done something fascinating from an engineering point-of-view,” says McManus. “Still, I wonder if I would prefer the traditional (Intel) architecture, so my investment in computing doesn’t have to change to accommodate this special board.”
McManus also points to Intel’s AVX chips, which are optimized for certain algorithms. Though different from Dragen, they underscore the chip giant’s ability to engineer their own specialized hardware. In addition, because Dragen is so new, Edico may face questions about the validity of its results.
“The downside of any of these tools is they don’t have a long record of publication,” says McManus. “Dragen is very impressive, you can get through your data really quickly, but the real test is when hundreds of scientific papers can cite its use and validate it properly.”
Poised to Grow
This validation process has already begun. In September 2014, Edico sold Dragen to Sequenom, which develops non-invasive prenatal testing. The two companies had collaborated on a study to confirm Dragen’s speed and accuracy. The processor is also being used by UC San Diego’s supercomputer center.
This is great news for Edico, validating their technology and generating cash flow. The company raised $10 million in July from Qualcomm Ventures, Axon Ventures and former Life Technologies CEO Greg Lucier. Van Rooyen hopes product sales will alleviate the need to seek more financing. They have an aggressive game plan.
“We’re targeting anyone that’s doing sequencing or is using sequencing data,” says van Rooyen. “The technology is ideal for hospitals, but our ultimate aim is to work directly with sequencing companies and integrate directly into a sequencer.”
While putting Dragen into sequencers would boost revenue, it also poses a regulatory hurdle, as the new machines would require FDA approval. Still, Edico is moving forward with confidence. In the past year, they have staffed up from 8 to 24 employees and plan to hire additional engineers and bioinformaticians. Van Rooyen is excited about their prospects.
“I’ve been through the mobile phone revolution and it was gratifying to be part of that. And to be part of the second big wave (genomics), which I believe is going to have an even greater impact than mobile phones, it’s very compelling.”
Information Technology Meets Life Sciences
Edico is one example of a larger intersection between life sciences and information technology. Google X is working with Novartis to commercialize a smart contact lens. But developing creative ways to handle huge data may be a particular sweet spot.
“A company like HLI (Human Longevity, Inc.), they probably generate somewhere around one petabyte per month,” says van Rooyen. “Your eyes water if you think about how much data that is.”
Van Rooyen brings a special skill set to the problem. An engineer by training, he started his career at the University of Pretoria, and later conducted telecommunications research for Sony. Ultimately, his entrepreneurial side took over. He co-founded Zyray Wireless, which was purchased by Broadcom after four years. He also co-founded ecoATM, which built mobile phone purchasing kiosks and was acquired by Outerwall in 2013.
In 2010, van Rooyen was working on a side project in his native South Africa, developing phone cameras as microscopes for remote health applications. Though the idea was sound, the camera could not detect tuberculosis, one of Africa’s most dire health problems. The real solution was sequencing.
“I had to learn about sequencing, and the difficulty of processing data,” says van Rooyen. “We got the idea to develop a chip that could replace all the servers, which led to Edico Genome.”