Yesterday, I wrote about a researcher “hacking” into Stephen Hawking’s mind with a new device called the iBrain. But for most scientists, hacking the brain doesn’t mean bypassing the physical body — it means making sense of the incredible glut of data presented by the human brain.
The Allen in the Allen Institute for Brain Science refers to Paul Allen, the billionaire cofounder of Microsoft whose $100 million in seed money helped the organization get off the ground in 2003. Last month, the Allen Institute held its first hackathon. The idea was to bring people who weren’t necessarily neuroscientists together to sift through the 6 million data points in the Human Brain Atlas, a massive database of gene expressions collected from donor brains and then put online for anyone to access.
(MORE: Paul Allen – The TIME 100)
In June, the Allen Institute created an API (application programming interface) to let users sort through the data equivalent of three human brains and several mouse brains. That’s a lot of information to process, which is why fields like computational neuroscience (which focuses on creating computer models of the brain) and neuroinformatics (which is more about finding ways to organize and analyze neuroscience data) are becoming indispensable.
So, who becomes an expert in something like neuroinformatics? People like Rich Stoner, a self-described “programmer academic” who put his knowledge of coding to good use after getting his PhD in bioengineering. The hackathon was organized to get people like him working to solve a thorny problem: How do you analyze and sort millions of data points representing something as complicated as the human brain?
“The scale of the data is beyond the capability of a single research lab or even an institution,” says Stoner. “Once you build a community around the tools, you get everyone speaking the same language and you can finally have a conversation.”
That conversation took place over five days in a single room, with the 30 “hackers” set up with 20 or so of the Allen Institute’s embedded software engineers. The environment, unlike some hackathons, had a loose, cooperative feel to it. There was no prize money, participants gave their presentations whenever they felt like they were ready to, and some of the projects seemed downright whimsical. According to Chinh Dang, the Allen Institute’s chief technology officer, one person composed a song with the brain data a while another made a valiant effort at creating an Angry Birds-style computer game.
“There was a great sense of community,” says Stoner. “The first night we were there, we coded until 6pm and then went to dinner, where we still talked about code. Then went back to the hotel and continued to talk and write code well into the evening.”
While most of Stoner’s research at the University of California, San Diego revolves around analyzing the post-mortem tissue of autistic children, his project at the hackathon wouldn’t look out of place in the App Store. He created a way to take data from the Allen Institute’s API and make it shareable in easy-t0-embed links; his hope is to eventually give instructors the tools to make instant educational materials for tablets and e-readers from databases that are constantly being changed.
The ironic thing about neuroinformatics is that while its reason for being is to sort through vast oceans of data, it also suffers from a lack of data. In other words, while three brains’ worth of material contains millions of data points, it’s still only three brains.
Human brains are not easy to come by, nor are they cheap to process. Last month when the freezer at the McLean Hospital broke down, researchers lost 150 brain samples including many from people suffering from autism, Alzheimer’s disease and other mental disorders. That was a huge blow to Stoner and his research on autism; his talents with neuroinformatics are of no use if there aren’t enough brains to study.
“There’s just not that much tissue available,” he says. Until processing becomes cheaper and databases grow, Stoner and his small but growing number of colleagues will have to make do with resources like the Human Brain Atlas, which is already the most impressive database of its kind. Hopefully as more students with coding skills move towards neuroscience and more donors give their brains to science, more databases like it will pop up in the future.