Expanse Bioinformatics has been awarded four new patents by the United States Patent and Trademark Office that cover data mining algorithms that the company has developed and plans to incorporate into a series of informatics products for the genetics market, the first of which it plans to launch within the next year or two.
The algorithms are designed to use a combination of genetic and behavioral information to suggest lifestyle modifications for patients that could help reduce their disease risks; suggest lists of possible treatments for diseases; find associations in social networks based on common genetic attributes; and make genetic-based product recommendations.
These algorithms are unbiased in the way they approach data, looking for any and all possible correlations between genetic markers, disease, and lifestyle data such as tobacco use and coffee consumption, according to Charles Eldering, the company's president and founder, In other words, they don't search for patterns in the data based on prior knowledge of relevant biomarkers or sequences, he explained to BioInform this week. Instead, "we let the data produce the associations … and correlations."
These algorithms also aren't intended to establish causality, he added, merely to help physicians and other users make the best decisions with available information. For example, if an oncologist has to choose between three FDA-approved treatments for a melanoma, rather than make a recommendation based on years of experience, he or she could use an Expanse product to analyze the patient's genetic information in the context of available information on melanoma and drugs to treat and then select the one treatment out of three that would be most effective in that instance, he said.
Expanse is currently working on its first product, a tool that makes personalized treatment recommendations for patients by looking for correlations between SNPs, diseases, and approved drugs. The tool will also be able to make behavior modification recommendations based on the integrated dataset to help individuals reduce their disease risks. Sometime in the next 12 to 18 months the company plans to package and sell the solution as software primarily to physicians and insurance companies, and further down the road plans to make it available as a service, Eldering said.
Aside from the obvious benefit to physicians and patients — more personalized treatment selection — insurance companies could also profit from the use of such a tool, he said, because they would be able to cut costs associated with paying for patients to test drive multiple ineffective treatments en route to finding the drugs that work best for them.
As part of their ongoing efforts to develop the yet-to-be named first product, Expanse is trying to forge partnerships with groups such as the Million Veteran Program (MVP), a project of the US Department of Veterans Affairs that aims to study how genes affect health. The project is collecting blood samples and health information from one million veteran volunteers to be stored anonymously for research into diseases such as diabetes and cancer as well as mental health conditions common to the military such as post-traumatic stress disorder.
PTSD is the first problem that Expanse hopes to apply its algorithms to, Eldering said. It "gives us a clear disease target," one that has the requisite genetic component and, assuming Expanse can partner with the MVP "[information] from a population that suffers [from the condition]." However once the solution is ready, it will be able to make treatment recommendations for a whole range of diseases, he said. Other potential non-disease application areas for the company's algorithms include obesity and weight loss, targets which Expanse will pursue in collaboration with yet-to-be determined partners who have the data it needs for such studies, he said.
Expanse is still mulling pricing options for its first product but it is considering a model where it is paid a percentage of profits that its customers take in from producing better outcomes at lower costs. Using the aforementioned example where an oncologist has to choose between three treatment options for melanoma, Eldering explained the pricing model this way; if in that scenario, "we can determine that the cost savings to the insurance company is $10,000 on average [per patient] based on the ability to make the genetic-based recommendation, then we want to shoot for a part of that whether that’s a third to forty percent" of the costs savings.
If this particular business model is adopted, the company would also apply it to other products its plans to develop for its portfolio. These include a social networking software that would allow users to find genetically similar patients — so-called SNP buddies. Customers of this tool — individuals suffering from a particular disease, for instance — could use the application to connect with other patients dealing with the same condition and form support groups through which they could share disease management tips, for example. This would also create a pool of potential study participants from a research standpoint. A third longer-term product would provide users with product recommendations that are tailored to their genetic make up, including things like cosmeceuticals, — cosmetic products that have biologically active ingredients — consumer electronics, and fashion choices.
In terms of setting a price point for a product recommendation application, for example, if the manufacturer of a $300 cosmeceutical finds that by making genetic-based recommendations they can increase sales and pricing of their products because customers are willing to pay more for a product that’s tailored to their genes, "we would expect to get a percentage of that lift," Eldering said.
Expanse Bioinformatics initially opened its doors in the early 2000s as a targeted advertising company called Expanse Networks which developed systems for figuring out consumer demographics based on their purchases and viewing habits. The company sold the IP underlying that business in 2004 and then re-launched in 2007 in Philadelphia with its sights set on developing products for the genomics market and a newly developed set of algorithms in tow.
Expanse sees IBM and Watson as its main competitor targeting the same sorts of customers and using an approach to analyzing data that’s along the same lines as its own. While its true that Expanse is a much smaller company — it currently has five employees, fewer resources, and is self-funded, though it plans to fundraise this year — Eldering believes that there is a share of the market to be had. The company intends to compete by focusing on very specific applications and demonstrating the benefits of using its solution to cut costs and improve care in the short term.