Opinion: Thinking Outside the Genome

In the last half century, the technology in genomics has provided us with a set of approaches initially as underappreciated as computers were in the early 1970s. “Exotic,” “finicky,” and “geeky” were terms used for mainframe computers that couldn’t even talk with each other. The same transformative technological advances that have turned computers into must-have personal accessories are inevitable for the nascent field of omics. Here are four ways in which omics will reshape the human experience.

DNA as a standard reference

What enables us to seamlessly share information across the Web is the core architecture called the TCP:IP address. It allows languages, international borders, and customs to all melt away, so that if I want, I can buy a rug on eBay from a dealer in Afghanistan. Similarly, the universal grids of latitude and longitude allow a coordinate reference standard that can tie together GPS, Yelp, and Zagat. The ability to generate a genomic signature for anyone at any time will at some point transcend its current informational utility to break through symptom-based descriptions of diseases and provide medicine with the standard reference needed to redefine the diagnosis process. And when existing symptom-based disease classifications are decoded into their actual omics-based subtypes, there will be a quantum leap in the ability of epidemiologists to track diseases. DNA will do for biomedical informatics what the TCP:IP address does for the Internet: it will allow information to be layered on citizens in ways that will make the promise of personalized medicine a reality.

Integrated omics models

We are about to undergo a phase shift, where a technology drives a fundamental change not just in what is known, but, more importantly, in how we think of ourselves.

Today we use omics technologies as isolated tools. DNA and RNA altered-component lists delineate altered nucleotides and amino acids characteristic of different diseases the way we would list damaged resistors and capacitors in a broken radio. Such lists are relatively easy to assemble. Soon, methylation and metabolite altered-component lists will also be readily accessible. The biggest changes in the next quarter century will not be in which one of these technologies displaces the others, but instead in how they are integrated. To build informative models of disease we will need to be able to construct predictive models that will rely on understanding how changes at the DNA and RNA level are reflected in protein and cellular processes. In a fashion similar to the way a number of information systems, such as the camera, the phone, the MP3 player, and TV, have been aggregated on smartphones, we should expect the layers of omics data to be used in fluid ways that offer much more in aggregate than in isolation.

How we will work together

Today most discoveries are made by scientist-clinicians who are funded to generate data, build a model or hypothesis, provide a validation of their idea, and then share the result as a paper in a peer-reviewed journal. This works well (and has for centuries) when the experiment is small and reasonably doable by a single laboratory. But as the scale of the data needed to make insights grows, as the algorithms to build the models become more complex, and as the necessary skills become diverse, the power of coordinated team approaches will grow. By analogy to physics, astronomy, and the writing of software, the benefits of dynamic teams sharing data and ideas in real time will multiply. The logical extension is to start considering a “commons” where omics data, projects, and models can be evolved in a shared manner. This will require a new set of rewards and incentives, but it will shift the current closed information system in biomedicine to a more open medical information infrastructure capable of the exponential learning associated with generative systems.

The democratization of medicine

As omics data and models accumulate, they will put pressure on existing guilds of experts. And as the information needed to diagnose diseases and track cures begins to be generated by “knowledge experts” who are not necessarily physicians, we should expect an activation of citizens empowered to help build better models of diseases. Patients in disease-advocacy organizations, such as the Love/Avon Army of Women or PatientsLikeMe, who have already been organized to track their symptoms and who can self-enroll in clinical trials, will recognize that they can move much further beyond the passive role of being “the sick.” In the near future, organized and active patients will start providing their own genetic samples for omics-based trials to determine whether they will respond to particular therapies. They will surely want to follow studies that they have enabled—and possibly even self-fund parts of future studies. And these involved patients will want to be educated about interpreting the outcomes of such trials. This establishment of a partnership between scientists, citizens, and physicians will be comparable to the democratization of publishing when the guild of editors and publishers partnered with the masses to launch Wikipedia. Individuals will realize that by using omics information, they can become the builders of the disease models needed for defining new medicines and deciding who should get which therapy. Patients will become copilots, jointly navigating the way to new therapies.

Stephen Friend is President of Sage Bionetworks, a foundation for building medical knowledge platforms where scientists and citizens coevolve models of disease.