Pharma-backed Toolkit to Speed Up Adoption of FAIR Data Principles
Blog Aug 07, 2019 | by Ruairi J Mackenzie, Science Writer for Technology Networks
Findability. Accessibility. Interoperability. Reusability. These are the four principles that together make up FAIR, a set of data principles that hopes to form a gold standard for scientific data management. First announced by a consortium of researchers in 2016, FAIR has gained traction among scientists and doctors frustrated by siloed data sources, confusing formats, and the wasted potential of huge volumes of analytical data.
One organization that is hoping to turn the flow of researchers and companies subscribing to FAIR into a flood is the Pistoia Alliance. The not-for-profit has announced plans, with the backing of major pharma companies, to develop a toolkit to help companies successfully implement FAIR. We spoke to Ian Harrow, a consultant at the Pistoia Alliance, to find out more.
Ruairi Mackenzie (RM): How will subscribing to FAIR data principles benefit pharma companies?
Ian Harrow (IH): Currently, much of the data created and used by life sciences companies are siloed in widely different formats and locations, making it difficult to retrieve, query, and share – essentially rendering it unusable. The combination of the ‘data deluge’ and life science data becoming recognized as a powerful corporate asset is driving the need for the adoption of the FAIR principles. At the same time, FAIR places specific emphasis on enhancing the ability of machines to automatically find and use data – a prerequisite for an industry increasingly looking towards AI and machine learning to support R&D. The toolkit we are building will help life sciences companies learn from each other as they undergo this culture shift together, supporting continued scientific progress.
RM: What form does your FAIR toolkit come in and how will companies utilise it?
IH: The FAIR toolkit will contain selected tools, best practices, training materials, use cases and methodology for change management, all of which will be delivered on a user-friendly, freely accessible website as a ‘wiki’, to be launched in Q4 2019. It will help organizations undertake their digital transformation, prepare for the Lab of the Future (LotF), and accelerate the application of AI and deep learning.
RM: Why is the Pistoia Alliance a suitable mediator between pharma companies in their goal of implementing FAIR?
IH: The Pistoia Alliance is a non-profit, neutral organisation, consisting of more than 150 members within the life sciences, biopharma, academia, publishing and technology industries, all of which have the same goal of enhancing scientific research and removing barriers to innovation in R&D. In its more than 10 years of existence, it has helped the industry work together on a variety of issues that lend themselves to pre-competitive collaboration, such as requirements gathering, standards setting, best practice assessment, non-proprietary data collection and sharing, and the like. FAIR is an approach towards data that our members are very keen to adopt, making this an excellent collaboration project for the alliance.
RM: How will FAIR principles feature in the Lab of the Future?
IH: The Lab of the Future (LoTF) will be digitally driven and will rely on automation. AI and machine learning systems will play a key role and they require high quality, consistent data to ‘learn’ from. Currently, as most data is not standardised or easily retrievable, it makes it difficult to impossible to ‘feed’ an algorithm without scientists spending large amounts of time data wrangling. In the LoTF, principles like FAIR are required to continue to modernise lab environments, help the industry continue to make breakthroughs, and shorten the development time of new research.
RM: Data is an increasingly valuable commodity for pharmaceutical companies. Why and how should they be encouraged to share this data with companies that are, in business terms, their competitors?
IH: FAIR is really about how we handle the data, its format, the standards that we apply to its collection, and so on, so there is no IP involved. In addition, large volumes of data that companies hold are actually pre-competitive, so again, it won’t affect businesses in regard to R&D and IP. Instead, by sharing data, it increases efficiency, prevents work being repeated unnecessarily, and allows one to ask questions that one could never ask before. Scientific breakthroughs increasingly happen at the boundaries of disciplines, and the crossover between science and other sectors – like technology – is growing. The problems that are being tackled require more data than any one organization could possibly hope to collect. To continue making breakthroughs, pharmaceutical companies will need to work closely with partners in other sectors, and collaboration and data sharing is key. Companies must recognize that not all data are proprietary, and non-competitive data could have great collective value. Ultimately, a more open approach to science benefits the most important stakeholder of them all – the patient.
Ian Harrow was speaking to Ruairi J Mackenzie, Science Writer for Technology Networks