In a world of interdisciplinary research, we need to make data freely available, says Katie Ember
Better Science through Better Data writing competition winner Katie Ember
Every Monday in the University of Edinburgh’s School of Chemistry, the Campbell group gather in Room 233 for a lab meeting. If you’re hosting the meeting, you bring cake. Or you forget and everyone pretends they’re not feeling a bit hungry and disappointed. Then, two scientists in the group present that month’s work.
Every Friday in the Centre for Regenerative Medicine, a fifteen minute cycle from the School of Chemistry, the Forbes group file into the first floor meeting room. After battling with the “motion-activated” lights, we all talk through what we’ve achieved that week.
The reason I go to two lab meetings in one week is because I’m attempting to detect liver damage using laser light. It’s multidisciplinary and it’s hard: requiring input from biologists, physicists and transplant surgeons from different institutes. The end result is that I spend about four hours each week not doing science but discussing it. Whilst this may seem like a strange way to do research, I cannot overstate how important it is.
Lab meetings are like open data on a tiny scale. We do experiments, share our results and get feedback from our colleagues. Not just from people within our field, but from people outside it. Because boundaries between disciplines are dissolving over time, research groups are becoming more diverse. A biologist will suggest I use a different type of cell; a physicist will say I need to try a different laser; a computer scientist will tell me something I pretend to understand. A problem I could’ve spent months agonising over alone is solved in moments.
Now, imagine that free flow of information on a global scale. We could troubleshoot, get inspiration and analyse data rapidly. After all, just because one institute has access to a particular resource, it doesn’t mean they can interpret all of the data they generate. Maybe I’ve gathered a huge amount of biochemical data from diseased liver tissue, but only a trained pathologist can to tell me about the exact stage of the disease: my data’s useless without the right person looking at it.
It is precisely because research is becoming more interdisciplinary that sharing data is more crucial than ever. Synchrotrons operated by physicists and engineers can generate a colossal amount of data from proteins, but it is biochemists who have the knowledge needed to solve their structures. Moreover, open data could see countries that are poor in certain resources become leaders in research by virtue of knowledge rather than funding. Science is a multi-step process and if an establishment lacks a mass spectrometer or an electron microscope, that shouldn’t hold them back from finishing a different piece of the puzzle.
So why is there reluctance over open data? Just as musicians would rather people didn’t download their songs for free, researchers can be unwilling to share years of work. But science is about discovery, not generating something from scratch. By taking a fluorescence image or a spectral measurement, I am not creating something new — I am simply revealing what was already there. That is the nature of science; it shouldn’t be owned by anyone.
In addition, open data will let us check that results haven’t been fabricated or are genuinely anomalous. It will enable us to ensure that the foundations of science we are building on are as solid as we like to believe.
Research is difficult enough as it is: there are barriers of discipline, language, terminology, funding and technology. We don’t need barriers to data as well.
Katie Ember is in her third year of a PhD in Optical Medical Imaging at Edinburgh University and is developing a way of sensing liver damage using laser light. She loves travelling, playing sports and writing. Follow her on Twitter for overenthusiastic tweets about scientific breakthroughs, space and the natural world.