Mendeley and Labfolder founders on how digital tools can revolutionize research
Victor Henning and Florian Hauer talk about the benefits of sharing raw data, the future of research communication, and their companies’ cooperation
By Alice Atkinson-Bonasio Posted on 3 April 2014
Last month, the global research collaboration platform Mendeley and the digital lab notebook Labfolder joined forces, enabling users to link publications to scientific raw data. Their cooperation not only improves the reproducibility of experiments, but also responds to a bigger trend in science and research towards becoming more open, connected and collaborative.
Dr. Victor Henning, CEO and Co-founder of Mendeley and VP Strategy at Elsevier, and Dr. Florian Hauer, COO and Co-founder of Labfolder, are both scientists and entrepreneurs. They believe that digital tools and platforms can simplify the scientific work process, connect scientists around the world and help them organize and share their data and literature more easily.
They recently had a conversation in about the benefits of sharing raw data, the future of research communication and scientific publishing, and how their companies' cooperation can advance these trends and address the needs of researchers today:[divider]
Q: In an interview with another outlet, you (Florian) said: "The future of research communication and data exchange will happen online." What does that mean?
Florian: I believe that digital tools and platforms improve the scientific process by making sharing easier and more democratic. In the previous century, only a privileged group was able to share information on paper. Digitalization has empowered the majority of people to communicate more data in a quicker way. Online tools have expanded the limits of research as more technical detail is shared and experiments are easier to reproduce. Storing data in the cloud also contributed to this modernization of traditional research.
Victor: There have been cases where scientists faked results, or drew premature conclusions which were then blown up in the media but couldn't be reproduced. From your experience as a scientist, how big of a problem do you consider the issue of reproducibility?
Florian: I think that most scientists are honest and document their findings correctly. The problem does not lie in the mindset of scientists but the possibilities of sharing data. When writing a journal article, results and procedures have to be heavily compressed. As technical detail is missing or gets summarized the reproducibility of experiments becomes more difficult. A large percentage of published papers cannot be reproduced because they do not contain enough detail or data. The research community as well as the private sector complains about this. We believe that the classical format simply does not offer enough technical detail.
Victor: That sounds great, but if I were to play devil's advocate, I would ask: Even if you manage to put all of this data out there, how will people know about it and discover it?
Florian: It is certainly easy to drown in the sea of data and not find what you are looking for. There are however more and more initiatives to restructure data and make it searchable. One example is the literature and data recommendation algorithm you use at Mendeley; it points scientists to the sources they are potentially interested in. Labfolder adds to this by enhancing the comparability of processes.
Victor: That is an interesting point. I once met the editor of Retraction Watch who told me that often people were not aware that a paper had been retracted. He suggested that we could introduce a feature to Mendeley where retracted articles would be flagged, and thus our millions of users would simultaneously be notified about the retraction. Would it be possible to extend this idea to raw data to inform people on the likelihood of a successful reproducibility?
Florian: Yes, I think that is a great idea! Working with raw data online would empower scientists to judge the trustworthiness of results. If data cannot be reproduced it could be flagged and retracted if necessary. This would eventually improve the quality of published data in the future. This is just one of the many possibilities that the sharing of scientific data in the cloud offers.
Q: Mendeley and Labfolder have recently launched a new cooperation. How does it address the needs of researchers and support new research trends?
Victor: Let me start by outlining the bigger ambition of Mendeley. When we created Mendeley, one of the ideas was to accumulate and crowdsource data into a central platform. We began working with bibliographic data because we wanted to provide a solution for efficient literature and document management. However, we anticipated including other kinds of data in the future as well. As our catalog and database grew, we decided to open an API to all of that data because we realized that researchers had far more needs and use cases than Mendeley could ever serve on its own. Our hope was that we could enable others to tap into our data and user base to fill these needs. Luckily, this is exactly what is happening today, and a great example of this is the integration with Labfolder.
Florian: Scientists and researchers read academic papers for two reasons: to find out more about a hypothesis or to gain knowledge on the method used. It might be irrelevant whether the experiment was done in a petri dish or a mouse and what matters to them is the method of accumulating the result. That is why it is interesting to connect literature with the processes described in it. Mendeley is offering the perfect tool for doing exactly that and it was very easy for us to implement their API into Labfolder. In this way it becomes possible to link the description of scientific processes with academic papers for the first time.
Victor: Do you think that in the future you could look at a paper in Mendeley and see that it has been cited in a process or "recipe" for an experiment in Labfolder? How would that work?
Florian: Mendeley's recommendation algorithm highlights the overlapping interests of readers. So if a person reads paper A and paper B, another person reading paper A might also be interested in paper B. Labfolder works similarly but focuses on processes instead of literature. In the future it might be possible for a person using process A to find out in what papers process A has also been applied. It would be interesting for that person to read these papers even if the subject might be completely irrelevant to their field of study. I could imagine that in the future one could look at scientific papers from two sides – the hypothesis but also the process used or the technical detail.
Of course it is always up to the user whether or not he makes the processes he uses public or not. One of the main purposes of a digital lab notebook is to secure intellectual property. This means that all content is set as private by default. However, during our work we observe that it is becoming increasingly attractive for scientists to share their data. Mendeley and Labfolder support and encourage this emerging trend.
Victor: One of the ideas behind Mendeley's recommendation engine was not only to recommend literature to people but also people to other people. The idea is that if you read the same literature as another person, you might also want to work together. The same goes for people working on similar experiments and data in Labfolder. In the future we could link both and connect people using a certain method on Labfolder to people reading about those methods on Mendeley. I think the idea of finding collaborators who perfectly complement each others' skills through Mendeley and Labfolder is a really exciting.
Q: What does the future of scientific publishing look like? Will there be "paperless" publishing?
Florian: We do not think the goal is to replace the way hypotheses are described in academic papers but to provide additional means for describing evidence. Academic papers of today only list the experiments that support the paper's hypothesis. The current format is not supporting the publication of negative results even though they can enrich the scientific process by providing information to other researchers. There is a trend that the publication and sharing of evidence is becoming more important. We would like to provide this additional format as a valuable resource for exchanging information. However the traditional academic paper will not be replaced by this additional offer.
Mendeley started out as a reference management tool and has grown to become a global research collaboration platform and academic database which connects academics, data and third party research tools. Mendeley's desktop, mobile and web applications help people to organize, share, and discover new research. Since its launch in 2009, Mendeley has grown to nearly 3 million users worldwide. The Mendeley API powers more than 300 third-party applications that are making science more social and open. In April 2013, Mendeley was acquired by Elsevier.
Labfolder is a digital documentation and planning tool for laboratory research, which was founded in 2013. It aims to expand the idea of a simple lab notebook into a platform where scientists can easily plan their experiments, document their data and collaborate with other scientists.[divider]
Victor Henning: I agree that it is not about replacing the academic paper but bringing it up to speed to what is possible with the help of modern technology. One project that demonstrates that Elsevier going in this direction is the Article of the Future which tries to make certain parts of a paper – particularly the data sets – more interactive by, for example, visualizing them in 3D. If somebody is using a tool like Labfolder, which preserves structured information, and this was integrated in the publishing process, it would make it much easier to bring formats like the Article of the Future to a greater number of people. One would not have to go through a painstaking process of reconstructing data scattered in different document and tools like Excel sheets, but would have access to structured data through Labfolder. This would simplify the publishing process and also make academic papers more interactive and accessible than before.
Q: Do you see a generational clash between researchers who work digitally and those working traditionally?
Florian Hauer: I do not see a clash. What I do see is a generation of digitally native scientists who have a naturally higher affinity to using online resources in labs. What we witness is more of a transition to this new way of working than a clash.
Victor Henning: It's true that more PhD students than professors use Mendeley, but that might simply be the case because there are more PhD students than professors in the world. I believe our user base is fairly representative of the demographics of academia – both in terms of disciplines and age. This is why I do not see much of a generational clash, although it is true that younger scientists who have not yet established a regular workflow routine are more open to experimenting with new tools. Someone who has worked with the same toolkit for the past 15 years is less likely to change. However, we have also seen cases where professors mandate their lab to adopt Mendeley.
Another interesting question might be whether younger researchers are more or less willing to share data than older ones. The argument has been made that older researchers have already achieved certain things in their career, like tenure, and that therefore their job is safe and they have to be less competitive. This enables them to share their data and engage with the community more freely than less established researchers. On the other hand younger researchers might be more open to working in communities and sharing data online. I guess we don't have enough data yet to answer this question!
Q: More and more scientists are also concerned with the structures in which science takes place, not only with content. You are both scientists and entrepreneurs. Are scientists increasingly turning into managers?
Florian Hauer: Pursuing an academic career – today or 20 years ago – automatically turns you into a manager. All you do is manage. This has not changed in the last decade. What is different today is that scientists are being more prepared for their management tasks during their education. One outcome of the rapprochement of science and entrepreneurship is that we get more and more scientific founders like us.
Victor Henning: When I was a PhD student, I was managing many different things such as grant applications, classes, and the supervision of master theses. This taught me a lot about team management and how to collaborate with other people. Obviously, as a professor you manage even larger teams.
Managing teams gets easier if you use the right tools that let you share information easily and effectively so that everyone in your team is always on the same page. Both Mendeley and Labfolder provide solutions to improve work processes within scientific teams. I think that technology shouldn't aim to replace interpersonal relationships, but it should aim to get rid of the unnecessary burden of administrative overhead.
Elsevier Connect Contributor
Alice Atkinson-Bonasio (@alicebonasio) is PR and Communications Manager for Mendeley, the company that created the Mendeley research collaboration platform and workflow tool and was acquired by Elsevier in April. She holds an MA in creative and media enterprises from the University of Warwick and is completing a PhD in online marketing at Bournemouth University. She is based in Mendeley's London office.