The transformative shift envisioned by FAIR pioneers is still very much still in progress. The standard reference tools, documents such as the FAIR Cookbook from the FAIRplus project and the FAIR Toolkit by the Pistoia Alliance, are exemplary. This report is designed to complement those other resources by documenting additional direct thoughts and insights from people who are leading FAIRification programmes on the ground.
Who contributed to this report?
We’re extremely grateful to the following people for their insights and contributions to this report:
- Lawrence Callahan, Chemist, Office of Health Informatics, Global Substance Registration System/Office of Health Informatics, Office of Chief Scientist, FDA
- Isabella Feierberg, Associate Principal Scientist, AstraZeneca
- Ben Gardener, Solution Architect – Knowledge Management, AstraZeneca
- Tom Plasterer, Director of Bioinformatics, Data Science & AI, BioPharmaceuticals R&D, AstraZeneca
- Ellen L. Berg, Chief Scientific Officer, Translational Biology, Eurofins Discovery
- Philippe Rocca-Serra, Group Coordinator and Associate Member of Faculty, University of Oxford e-Research Centre
- Martin Romacker, Senior Principal Scientist in Scientific Solution Delivery and Architecture, Roche
- Andrea Splendiani, Director of Data Strategy, Novartis
Who is this report for?
Those in data architecture roles including, but not limited to, those in data analysis or data science focuses; as well as data repository managers, policymakers, strategic leaders, project coordinators and researchers who need to ensure that their data is reusable and publishable would benefit from reading this report.
Contents
Prologue: Advice for starting a FAIR journey
Chapter 1: Why does FAIR matter?
- Why FAIR implementation is important in pharmaceutical R&D
- An outline of the cultural and technical challenges
- A synopsis of the leading resources available on FAIR standards and implementation, with a short explanation of what each offers
Chapter 2: People and processes
- How many people are needed to FAIRify the data landscape?
- Collaborating and convincing – what management is required to implement change?
Chapter 3: Moving from an application centric perspective to a data-centric one
- The problem with an application-centric approach
- What’s required to become data-centric?
- Academic implementation vs pharmaceutical implementation
- Do data lakes help or hinder creating FAIR data?
Chapter 4: The challenges of enterprise-level FAIR implementation
- Why is FAIR harder at scale?
- The cost of legacy data
- Changing attitudes towards FAIR on this scale
Chapter 5: Evaluating FAIRness at scale
- Does everyone involved in the data lifecycleunderstand how deep FAIR needs to go?
- The FAIR metrics
- Maturity models for FAIR – how do we score ourprogress? What models are out there?
Chapter 6: Can all graphs be FAIR?
- Knowledge graphs as the killer use case for FAIRdata
- Label property graphs versus RDF graphs
- Graph trimming
- Applying multiple perspectives to datasets
Chapter 7: The impact of AI and ML on FAIR data use
- A brief look at what the FAIR principles can offer for advanced analytic model building
- Where are AI and ML making most notable strides in pharmaceutical R&D?
Chapter 8: How far along are different sectors in adopting FAIR standards?
- A contributor each from pharmaceutical R&D, a contract research organisation (CRO) and a regulatory body (the FDA) discuss what they consider “success” in terms of FAIRification.
Epilogue: Future Needs
Download