The Era of Big Data Analytics

February 8, 2011 | Russell Transcript | Toward the end of 2010, two systems biology (SB) pioneers, GNS Healthcare (formerly Gene Network Sciences) and Selventa (formerly Genstruct), underwent name changes and took major steps to reposition and grow their businesses. Time will judge the effectiveness of the moves. Today, the systems biology label seems mostly forgotten while its tenets live on, and perhaps that was to be expected.

The two companies are very different in most ways but alike in one fundamental aspect. Both believe in, and are committed to data-driven, computer-powered modeling and simulation to drive drug discovery and health care delivery decisions. If that refrain sounds familiar, it should; many companies have issued that call but none have turned it into broad commercial success. Recently, the pharmaceutical industry has seemed a little tone deaf on the topic, faced as it is with bigger fish to fry.

The GNS Journey

For me, the question has always been timing. “The birth and growth of high throughput technologies combined with super computing gave us the tools to systematically unravel the cause and effect relations responsible for disease progression and drug response. That was the concept and driving force behind GNS from day one,” says Colin Hill, CEO and co-founder. “It just took much longer than expected.”

“We’re finally seeing the generation of data from many drug development projects on a scale such that you can really do systems biology. We see major breakthroughs coming from the application of these technologies in areas like autoimmune diseases, multiple sclerosis where we have a big effort, and very much in oncology. I think we’re going to see massive breakthroughs in oncology over the next 5-to-10 years and we expect to see our technology at the core of a lot of these breakthroughs,” says Hill.

Now a decade old, GNS’ secret sauce has always been its mastery of mathematics, statistics, and machine-learning techniques embedded in the GNS REFS (reverse engineering and forward simulation) platform (see, “Tailoring Medicine with Supercomputers,” Bio•IT World, Mar 2010). Among other things, the REFS engine can infer underlying rules without prior knowledge. While turning data into dollars has proved difficult, “we kept developing and refining the tools and the platform,” Hill says.

Indeed the growing flood of data in virtually all industries prompted Hill and his colleague to look for other industries in which to also apply the REFS platform. Two years ago GNS spun out FINA Technologies to apply the REFS technology in financial markets. Just this past November, Hill launched Via Science, a holding company and incubator. Gene Network Sciences was renamed GNS Healthcare and along with FINA was brought under the Via Science umbrella. The broad idea is Via Science will seek out new data-rich markets for the REFS technology and spin out companies to exploit those opportunities.

“We think the best way to do that isn’t to attempt do that all within one company but to essentially incubate the launch of new companies that are led by domain experts who come from those areas but also have quantitative backgrounds,” says Hill who is also CEO of Via Sciences. “Hopefully we can rapidly and cheaply determine if the REFS platform and related technologies have differentiating value compared to the current state of the art method of crunching data in these verticals.”

Consistent with the goal of finding new REFS targets of opportunity, GNS has expanded its offerings. Biotech and pharma still comprise a major customer set with GNS plowing through ’omic and clinical data to help identify mechanisms of action, identify useful biomarkers, targets, responder populations and the like. But the company is also offering solutions to payers and providers around risk characterization, formulary management, and a range of other tasks.

Despite wearing two CEO hats, Hill says, “My personal focus is still very much within the health care realm and I actually think the era of systems biology is finally coming. Whether it comes under a different name or not, the supercomputing hardware and machine learning platforms such as ours are finally advanced to a point where we can really do system biology.”

Selventa Gets Personal

Although its technology is different, Genstruct also spent nearly a decade building a predictive platform able to make sense of large datasets and has successfully conducted on the order of 60 engagements (see, “Patience, Persistence, and Payoff,” Bio•IT World, May 2008). In late November the company changed its name to Selventa. A week later it named a new CEO, David de Graaf, who had joined the company in June as CSO, having been brought in from pharma (Boehringer-Ingelheim, Pfizer, and Astra-Zeneca) to help drive commercialization.

de Graaf echoes Hill’s idea thoughts on timing.  “I think we are now beyond technology stories. When we started talking about this space 7-10 years ago, a lot of it was about just validating the need to look at complex biology and using computers for these purposes. Now, that has become somewhat old hat. People not only understand the need but also may have some piece of it in place,” says de Graaf.

The Genstruct Technology Platform—it’s still called that—combined a proprietary “language” to capture biological concepts and make them ‘executable’ and computation engine. In typical, perhaps necessary, fashion the company did many proof-of-concept projects around hypothesis generation and biomarker identification. Like GNS, it too claims to unravel causal, predictive relationships.

de Graaf says the early Genstruct pitch was around exploring difficult and deep biology with clients and then pursuing fee-for-service projects to develop relationships the might lead to broader engagements. Today, Selventa has become a personalized medicine partner with defined solutions areas and a distinctly commercial focus.

“We can take forward decisions that tend to happen fairly late such as in Phase 2B and Phase 3 with regards to response and population size and understanding of mechanism of action and bring those decisions much earlier, in some cases all the way to early discovery,” says de Graaf.  Strategic partnerships, fee-for-service collaboration, and technology licensing all part of Selventa’s offering portfolio.

de Graaf says fee-for-service projects are still a big part of the company’s business, but they now may include technology licensing components providing a way for clients “take our results and rather than just getting a presentation at the end to bring models in(house) and then to grow and develop those models of disease themselves.” (GNS too is making a greater effort to license its technology directly to clients. Both companies say doing this is a measure of the technology maturity and improved ease-of-use.)

Long term strategic relationships, says de Graaf, is where “we share in the upside and obviously there are reduced upfront costs in these types of structures. We help clients with patient stratification and development of diagnostics whether those are therapeutic diagnostics or companion diagnostics as well as the positioning of their portfolio.”

Unsurprisingly, the new CEO has a long to-do list including for example: making deeper inroads in academia and opening the company’s BEL (biological expression language) Framework to get more people validating the platform; crafting still more granular solutions (e.g. predictive toxicology and patient stratification); and streamlining the fee-for-service component.

It seems clear the Big Data Era is upon us. The future progress of GNS Healthcare (and Via Science) and Selventa will be indicators of when the era of Big Data Analytics arrives. 

This article also appeared in the January-February 2011 issue of Bio-IT World Magazine. Subscriptions are free for qualifying individuals. Apply today.