Crowdsourced science could be future of nutritionCrowdsourced science could be future of nutrition

Ask people what a “scientist” looks like, and they’ll likely describe a bespectacled, PhD-holding man in a long white coat. He comes up with a hypothesis and works up a grant request to plead for a slice of the ever-shrinking pie of public research dollars. If he gets it, he convenes a small, carefully selected group of patients to test the theory that, for instance, nutrient A or pharmaceutical B helps prevent or quell condition C.

Jessica Richman sees a different image. “We live in a society where a select few people are anointed question-askers and others are not,” says Richman, co-founder of the San-Francisco-based citizen science start-up uBiome. “I see a future where all of us can generate hypotheses and set research agendas, where anecdotes can converge into data to create meaningful research, and where everyone who wants to be, can be a scientist.”

Those words make some traditionally-trained scientists bristle, but in a sense, that future is already here.

Thanks to the explosion of social media, the plummeting cost of high-throughput gene sequencing, and the growing willingness of consumers to share their health information online, dozens of “citizen” or “DIY” science efforts have emerged to study everything from how the gut bacteria of a vegan differs from that of a meat-eater to what supplements work best for alleviating Crohn’s disease or what genes fuel Parkinson’s. The movement was pioneered by health-technology companies like 23andMe,
which—since 2007—has provided a detailed genetic analysis of customers’ spit in exchange for a $99 fee. The customer gets information about him or herself that could potentially help them achieve better health. In return, 23andMe gets to add their anonymized analysis to a massive database for further research. Now a slew of other companies, patient advocacy groups, and even academic researchers are jumping on the citizen science bandwagon, inviting consumers to provide not only real-world biological samples and anecdotes but also ideas on what research should be done with them and money to carry it out.

Combine enormous data sets with crowdsourced funding, plus a new pipeline of ideas from the people who care the most about the results (the patients), and progress toward personalized treatments could be greatly expedited, says Larry Smarr, founding director of the California Institute for Telecommunications and Information Technology (Calit2).

“Medical science has relied mostly on people within the medical establishment,” says Smarr, who personally used a host of DIY medical tests, including DNA, stool, and blood tests, to diagnose himself with Crohn’s disease. “Where we are going now is very different. It’s a huge opportunity to change medicine as we know it.”

The power of poo

If there is one health subject sure to get the public interested in participating in DIY research, that subject is poo, says anthropologist and microbiome researcher Jeff Leach, co-founder of the American Gut Project, “the world’s largest open-source science project to understand the microbial diversity of the human gut.”

“Everybody takes a crap in the morning. That is something we all can wrap our head around,” says Leach. Once people realize that 75% of that stool sample is bacteria, and that there are tools that can enable them to compare their bugs with those of other people, they will be empowered to take steps toward better health, he says. “In the next five years, I imagine that people will be able to get their gut microbiota sequenced on a monthly basis for, say, $19.99, and use that information to nudge their microbiota in the right direction with lifestyle and diet changes.”

But first, they have to know what a “healthy” microbiota—the collection of microbes living in and on us—looks like, and that takes massive amounts of data. That’s where American Gut comes in.

Prior to American Gut, the largest effort to map the human bacterial ecosystem—the National Institutes of Health Human Microbiome Project—looked at about 230 individuals. “Two-hundred and thirty healthy people from college age towns is not what the world looks like. I want to know what the couch potato looks like, or the ultramarathoner looks like, or the guy on the Atkins diet looks like,” Leach says.

Getting funding for such a broad project was impossible via conventional means. So Leach teamed up with University of Colorado-Boulder professor Rob Knight, PhD—among the most respected scientists in the field of microbiome research—to launch American Gut in November 2012. The project has since collected nearly 10,000 samples (oral, skin, and gut) from 70 countries, which are run through a machine that sequences the DNA of the microorganisms present. It has raised nearly $1 million, both from the sample donors (who pay $99 for a print out of their resident microbes) and from companies like Second Genome (an R&D company specializing in “microbiome modulators”) who sponsor kits for people who want to donate their sample but can’t pay the $99.

The project has yet to release findings in a formal peer-reviewed research paper, but it periodically posts updates on general trends emerging from the data (people who exercise more and eat a greater variety of plants have a more diverse microbiota; people’s inner-flora changes with the seasons; the gut bacteria of a vegan looks very different than that of a meat eater). It stresses that it does not provide “actionable health information to its donors” and that the greatest thing they get for their $99 bucks is a chance to contribute to important science.

In keeping with the open-access philosophy that often goes with citizen science, American Gut promises to release the raw data freely to the public “as soon as it comes off the sequencer.” “It invites perspective from people who maybe couldn’t get a grant in the traditional system but might have something interesting to say about the data,” says Knight. Ultimately, he believes the data could also be used to help companies make probiotic cocktails that really work. “There is a huge amount of intuition in the field at the moment—you see whole aisles of probiotics at the health food store with no particular basis for choosing one organism over another. Replacing that intuition with empirical data would be very helpful.”

uBiome—a similar, for-profit version of American Gut—takes a slightly different approach. “Our mission is very consumer-focused,” says Richman. A charismatic young Stanford grad with degrees in economics and science, technology and society, she founded the company in the fall of 2012 to bring the power of microbiome sequencing to the people. In 10 weeks she raised $350,000 via Indiegogo.com.

To date, uBiome has collected 7,000 samples from 4,000 people, who pay $89 for a detailed peek at what bugs they host and what that means for their health. Right now, she concedes, most of the information is “just interesting,” but “actionable results” are not far off.  “Within five years we will be able to say here is what your gut was like before you took antibiotics, and here is how to get it back there. Or, if you are going to be taking chemotherapy, here is what you should take with it to have the least side effects on your gut bacteria.”

uBiome already encourages its clients to come up with their own experiments, looking at their gut microbiome before and after a certain intervention—like an added supplement or diet change—and it is already working with several CPGs on research using its vast citizen-fueled database to help them develop things like probiotic toothpaste, skin cream, or shampoo. “If someone is making a targeted probiotic supplement—specifically for women, or vegans, for instance—we  can do a large study to help them discover what is actually in the healthy gut of a certain population,” Richman says.

Beyond the gut

Gut bacteria aside, citizen science is making its way into a host of other realms to answer an array of intriguing questions. Despite 23andMe’s recent run-ins with the FDA (the agency recently demanded that it discontinue its popular direct-to-consumer Personal Genome Service), the company has amassed genetic and lifestyle information on more than 650,000 people, and used it to support more than 230 studies, some by their own researchers, some by outside collaborators. Some findings have been trivial—one study identified the genes that make certain people’s urine smell like asparagus after they eat it; another identified the genes that make some people sneeze when they look at the sun—but others could prove valuable. For instance, one recent study, using survey and genetic results from a whopping 15,072 23andMe customers, was able to identify 11 genetic variants that boost risk of asthma/hay fever. 23andMe also boasts the largest single cohort of genotyped Parkinson’s patients in the world, which has enabled its researchers to pinpoint gene variants that may predispose people to the disease.

This spring, the non-profit research foundation Sage Bionetworks reached out to 23andMe and other big data sources to help Sage launch the Resilience Project, a citizen science effort aimed at pinpointing “unexpected heroes” who were genetically destined to get a grave childhood disease but never did. “We think the way to find ways to prevent disease is not to study what everyone studies—which is people who get sick—but to flag people who should have gotten sick and didn’t,” says Sage co-founder Stephen Friend, MD, PhD.  So far, the Resilience Project has looked at genetic records and health surveys of more than 590,000 people. Now, it is appealing to the public directly via its website to donate DNA and fill out a health questionnaire to determine if they may be an unexpected hero. Once the project identifies a cohort of people who fit the bill, it will take the next step, looking at what those heroes may have in common. “We are looking for the protective genes and protective factors that keep people from getting sick,” says Friend.

His organization also develops open-source technological tools to help patient advocacy groups and research teams launch scientific initiatives on their own. One such project will invite 600,000 users of the free mobile phone iSleeping app to join a study that will monitor and report how different sleep medications modulate sleep patterns. Another will use data from self-monitoring devices—like step-counters and calorie-countering phone apps—to look at how lifestyle changes impact Type 2 diabetes outcomes.

Meanwhile, PatientsLikeMe, a patient network with 250,000 members suffering an array of diseases, is offering up the real-world anecdotes and shared data of its members to help biotech collaborator Genentech come up with better solutions. And Crohnology.com (an information-sharing site for those with Crohn’s disease or colitis) is aggregating member’s personal anecdotes about the medications and supplements they take into databases which are then shared with the group to help members make decisions. “You are asking the people who actually have the disease in the wild, where it counts—not in some artificial study—what is working and what is not,” says Smarr.

Skeptics of the citizen science movement are quick to point out that some patient-driven science projects lack the scientific rigor that something like a double-blind, randomized controlled trial delivers. Others have expressed concern that people may be jeopardizing their own medical privacy when they sign on to donate a fecal sample, or swab of saliva, and have it genetically sequenced for the world to see. uBiome, for example, was highly criticized for soliciting funds via Indiegogo before clearing its idea with an Institutional Review Board to assure its research was carried out humanely and ethically.

Smarr says some of the complaints are certainly valid. “Is citizen science as well controlled as laboratory science? Probably not,” he says. But because the data is so big, “you can probably still come up with some pretty accurate results.” He sees citizen science not as a replacement for laboratory science, but as a critical adjunct that enables the public to offer its input, scientists to more openly share their findings, and companies working to make better health products do so more quickly.

His advice to companies in the nutrition industry? Start paying attention, and start thinking of ways you can use citizen science to engage your potential customers to help you make products that work better. “My guess would be that in 10 years, any company that is not crowdsourcing data in some way will be out of business because it will not be competitive,” says Smarr.