Using the internet for open science
A few days ago, PLoS hosted a talk by Michael Nielsen at their San Francisco offices. Nielsen, author of the book “Reinventing discovery”, due late this year from Princeton Press, is a strong-voiced proponent of the need for a change in the way we share data.
The Polymath project, his opening story, is one of the best examples of how and why open science works. Tim Gowers, a Fields medalist, posted a famous mathematical problem on his blog, an open invitation to anyone interested to try their hand at solving it. For the first 70 hours, nothing happened. Then a math professor left a comment, quickly followed by a high school teacher, another Fields medalist and so on. In the span of 37 days, over 800 comments collectively solved the problem. How many conferences and scientific papers, peer reviews boards and editorial revisions would it have taken to even get these diverse minds thinking together in the same space? Nielsen describes it as the difference between “driving and pushing your car”.
“Using the internet to build tools that amplify our collective intelligence is critical”, as Nielsen says. Yet apart from anecdotal successes (and just as many failures), the scientific community is yet to use the internet to effectively change the way we communicate. A previous post on our blog describes John Wilbanks from Science Commons as saying, “We digitized science, but we didn’t make it digital.”
Given the ample potential of the internet, why is it so difficult to use it for science? One of the reasons Nielsen points out is the absence of metrics that recognize or reward collaboration. “It is still more worthwhile for an early career academic to spend their time writing a mediocre paper and getting it published than to be writing a really good encyclopedia article about quantum computing on a Wiki.”
If scientists are to share data and collaborate in real-time, information sharing needs to be recognized as an essential aspect of scientific discovery. Given the right incentives, a change in scientific mind-sets can produce remarkable results- the Human Genome Project and the Alzheimer’s Disease Neuroimagine Initiative (ADNI) are ample proof of the power of open science.
And if the many emerging successes of collaborative research aren’t proof enough, Nielsen reminds us that we are really the beneficiaries of a previous “open science” revolution. A few centuries ago, the idea of publishing a paper was unheard of. Over 200 years ago, Galileo was the first person to observe the rings of Saturn through his telescope. Like any conscientious scientist, he wrote up his discoveries in a scientific report and mailed it out- as an anagram. Sending reports of new discoveries as anagrams wasn’t unusual for the time- Da Vinci, Newton, Hooke and several others were all guilty of it. Publishing as it stands today changed that culture of scientific secrecy (and even sabotage).
The shift from cloak-and-dagger science to more rational information exchange was largely due to the printing press. Providing the tools for communication helped the scientific community establish the platforms we now think of as familiar: Journals, peer review processes, conferences and societies.
The World-Wide web, our most powerful tool for a second paradigm shift in the way we exchange information is well enough established already. We’re using it to do everything from buying shoes to stalking an ex- or watching the end of Optimus Prime.
We only need now to leverage these new technologies to scientific discovery. Portals for the exchange of ideas and integrative data sharing are laughably rare. Where are the websites with suggestions like: “Related genes that people searched for-“, or “People whose research interests match yours”? Discussion boards for research problems or reliable reviews of laboratory products can be just as hard to find. Platforms like NextBio (which does make ‘personalized’ recommendations!) focus on integrating data from diverse lines of research, and could not exist without the fundamental principles of open science and data sharing.SHARE