Publishing in the 21st century: a blog for researchers
What do you do with your old research data? When I finished my final postdoctoral position a few years ago I had several gigabytes of data stored on various portable hard drives scattered around the lab. A large portion of this data consisted of high resolution confocal image stacks, 3D projections and single images which could demand up to 150MB in disc space each. Although we had also been storing our confocal data in a custom built lab database, much of it was still unpublished and the data was only visible to lab members. Even though I had published papers based on my data, the actual confocal image stacks and supplementary files were not in the public domain.
We were not the only lab to face this problem, groups that we collaborated with also had libraries containing thousands of confocal images stashed on hard drives that they couldn’t share easily. The easiest way for me to explore another group’s image data was to visit them and spend a few days in front of one of their computers; not ideal when they were based 4,000 miles away.
Since then, the data sharing/storing landscape has changed. Most funding bodies now require researchers to deposit data in institutional repositories and there are also cloud based storage options such as Dryad. My current favourite for storing and sharing data is Figshare, where you can deposit figures, media, datasets and more. Each item is given a unique doi, so it can be cited using traditional methods and you can upload as much as you like as long as you make it public (plus you can have 1GB of private space). I’ve been uploading files and playing with Figshare for a several months now and find it to be a powerful platform for disseminating research and finding hidden nuggets of science. It includes metrics for each item so you can see how many people have viewed and shared your file; the most popular file on my profile is a short video of developing neurons (frozen in time) in the fly brain.
I’m no longer doing research and hate to think of all of my data sitting on a dusty shelf in the lab and gradually being sucked into the black hole of research obscurity. Therefore I am now in the process of uploading the majority of my PhD and postdoctoral work to Figshare along with descriptions of the data. I’m starting with the confocal image stacks from my PhD which investigated the adult structure and development of the central complex region (the centre for locomotion control) in the Drosophila brain.
For the project, I used the handy Drosophila genetic toolbox to reveal different neuron expression patterns in the central complex, a region of the brain comprised of approximately 1,000 neurons, and a combination of genetic-driven fluorescence tools and antibody staining. In order to investigate the development of the structure, fly brains had to be extracted, stained and imaged every two hours throughout development (metamorphosis in this case – it lasts ~100 hours) using a variety of different antibody markers and expression patterns. As you can imagine, this generated a sizeable dataset to analyse – over 1,500 image stacks and considerable time in a dark cupboard alone with the confocal microscope. I’m now uploading it piece by piece to my Figshare account. By releasing the data this way, several years of worth of work can be shared without being lost and people can build on it as they see fit. The data will complement our previous publications and hopefully the images will be a useful resource for other researchers when designing future studies.
Public platforms such as Figshare are also effective tools for posting negative results which would not normally be accepted via traditional publishing channels; this is also being addressed by the open science journal F1000 Research, who are offering free publication for articles reporting negative results until the end of August. On my Twitter feed a few months ago someone posted that approximately 70% of scientific research goes unpublished; hopefully the existence of sites such as Figshare, F1000 and Dryad will encourage more researchers and ex-researchers to share their data whether it’s positive, negative or vintage.
Posted by Jo Young