Renata Osika

The growing need for open science

By Emmanuel Mongin and Renata Osika

For decades, public funding for science has been provided on the premise that a better understanding of the fundamental sciences will ultimately lead to a more prosperous and healthier society. As a result of a seminal essay by V Bush — Science: the Endless Frontier (1945) — industrialized economies have benefitted from successful, diverse programs that support basic science and are aimed at generating answers to the challenges identified by societies.

Faced with increasing challenges associated with issues such as the rise of non-communicable diseases, energy and water scarcity and global environmental issues, govern- ments and societies have embraced science as a way to solve large social problems. This is reflected in new projects created to answer big science questions through large-scale scientific undertakings such as the human genome project or climate change initiatives. This has fuelled tremendous growth in scientific outputs while opening new doors. It also presents new challenges and comes with the increased burden of data management.

Big science questions are extremely complex and require large-scale collaborations. In response, "open science" emerged as a means for collaboration, data sharing, and transparency. Although open science has always been part of research practices, large-scale multidisciplinary research projects have popularized such practices and given them appeal.

The human genome project was undoubtedly a successful open science project, benefitting society at large beyond short-term economic gains. Its magnitude and importance triggered new incentives for the scientific community (both public and private) to find novel collaborative methods. While open science doesn't necessarily overturn the methods of the traditional silo culture of science, it represents a step towards making data inter-operational and accessible to help answer complex questions.

The need for open science is more urgent than ever before. With many data-heavy research projects still in a data production stage, open science offers mechanisms such as alternative licensing to enable data sharing and exchange, and also various open-source software approaches for data analysis. This paradigm shift promotes a collective mindset, where both publicly funded fundamental research and the private sector research share data to more efficiently produce, analyze, and apply scientific knowledge.

Although the private sector remains the largest funder of R&D, fundamental research is largely publicly funded. Pressure has increased to demonstrate how tax dollars are used for societal benefits. Large multidisciplinary science projects need the integrated data, some would argue, from all available sources. Proprietary science is sometimes seen as a hindrance.

We recognize that both approaches are necessary: proprietary approaches promise to protect the potential return on investment; open approaches favor the development of common standards and tools for pre-competitive research. Open science is often mistaken as a free-for-all approach, when in fact it offers alternative licensing mechanisms that protect authors' intellectual property, while facilitating and permitting data sharing without restrictions of the traditional copyright paradigm.

In this post-genomic age, the amount of data available tremendously increases the amount of potential drug targets available for study. Testing of all potential targets is far beyond the reach of any one publicly funded project or the pharmaceutical industry alone. To protect their investments, studies done by the pharmaceutical industry tend to stay within their laboratory walls, with significant overlaps among firms. Unless they make it to market, the generated data and knowledge cannot be applied to create social returns.

The Structural Genomics Consortium (supported by universities, private sector, and governments including Canada and Ontario) proposes new solutions to push the limits of pre-competitive research. Its main goal is to determine and put in the public domain the three dimensional structures of medically relevant human proteins. Such projects push the limits of pre-competitive research by making freely accessible data relevant to drug design available to the scientific community. The end result is freed-up R&D capital for investment in drug targets likely to be the most successful.

An open approach to science is also an effective way to combine and analyze data from complex and interdisciplinary research agendas such as climate change. Understanding its causes and effects comes from many scientific fields, from physical sciences to chemistry. Taken alone, any one of them produces variables that provide limited utility, unless analyzed in a global climate system.

Due to its multidisciplinary nature, climate science has been a very open field with datasets coming from various types of organizations and countries. Large scientific collaborative platforms based on openness of data, software and analysis may prove effective for leveraging not only the global web of expertise and data, but to entrust local and practical solutions to a range of community-level climate related challenges. A tendency towards joint initiatives can be observed with agencies such as NASA or the European Space Agency, with supporting policy, computational frameworks, and public access.

However, unlike the human genome project, climate change is a political hot potato. Geo-political considerations (economic, security, etc) that underpin climate change debates may in fact be the major road block for openness and data sharing within the climate science milieu. Nonetheless, more openness could help reestablish public trust, while continuing to support merit-based research excellence reinforced with open collaborations, shared data and some form of open peer review.

Open science is more than data sharing on the world wide web. The increase in inter-disciplinary projects tackling large social challenges may help break the silos and protectionism of the copyright and patent mentality. In order to leverage the promise of the computational prowess of the 21st century, stakeholders of data rich science projects will need to come together in integrating various knowledge sources. Developing capacity to bring fundamental research data to those who implement and apply it is the necessary prerequisite before society reaps the benefits of both public and private investments.

Renata Osika has been a senior management consultant in areas of business transformation and systems integration and has background in computer science. Emmanuel Mongin has worked in the field of genomics. They are currently employed at the Council of Canadian Academies. The views expressed in this article are solely those of the authors.