The development of a ‘Findable, Accessible, Interoperable and Reusable’ (FAIR) data approach has been an exciting step towards an information-centric approach for the pharmaceutical industry. The potential benefits of this alternative approach to data management and accessibility could significantly support pharmaceutical R&D and drug development. However, adopting this new approach could be challenging for some organisations and may not be as popular as anticipated. 

For daily articles on the latest pharma trends and innovations, as well as interviews with leading experts and in-depth industry White Papers, subscribe to PharmaFeatures.com.

Introduction to FAIR data

Two of the main problems limiting drug development and clinical research are the accessibility and management of data. There are vast genetic databases, medical records and scientific archives which have the potential to support the pharmaceutical industry but have only just begun to be utilised efficiently.

FAIR data is a recently developed concept built on a number of principles which aim to support drug discovery through good data management, which are as follows:

• Findable – Data are richly described by metadata and have a unique and persistent identifier

• Accessible – Data and corresponding metadata are understandable to humans and machines, and accessible through defined protocols

• Interoperable – Data and corresponding metadata use formal and accessible knowledge representation to guarantee reuse

• Reusable – Metadata accurately describe the provenance and usage license for the data

Findable is one of the most important principles for FAIR data and is built on the premise that data must be of high quality when it is inputted into the database. In other words, up until its most recent update, data must be meta-tagged to ensure the record of the data is up-to-date and accurate. This includes “information on the author, those that have access to the data, who has modified it and a whole series of characteristics such as category, security classification and file type”. 

This metadata is critical for intelligent databases to filter through vast datasets to retrieve relevant records.

The ultimate goal of FAIR data is to help lower R&D costs, reduce drug development timelines and prevent late-stage failures. Extensive analysis of data across multiple platforms could help researchers to better understand a disease and potentially modify preclinical models to better assess drug pharmacokinetics in a biological system. 

Challenges

The principles for FAIR data helps users to get the most out of the available data, however implementation in the pharmaceutical industry has been slow for two main reasons, according to a publication.

Firstly, the implementation of FAIR data is not necessarily pre-determined or standardised. FAIRification is a long-term solution of how data is created and used within an organisation, however this concept continues to change as our knowledge of processes (biological and digital) evolves. 

This was a point emphasised in the aforementioned publication which states that a critical part of FAIRification is the mapping of data according to a semantic model. “This model describes the meaning and relevance of each data point, and its relationships to other data in the context of a knowledge domain. Gaps in our understanding of biology make this mapping a moving target. New information fills in what we don’t know and changes what we thought we did know.”

The second reason contributing to slow implementation is the potential cultural implications. For those in the pharmaceutical industry who have always worked in information silos, shifting to sharing and reusing data may be considered counterintuitive. Hence, adopting the FAIR principles may be a challenging and arduous process for many organisations. 

Other challenges raised include the need for “clear data licensing, how to proceed and what license to suggest/apply to data from a wide variety of sources (governments, agencies, states, universities, investigators, etc)” and how the need for user identification for access to platforms may introduce a new barrier to data access. 

Importance of FAIR data

The implementation of FAIRification across the pharma industry has potential to support everything from R&D to drug development. According to a recent Nature article, the following stakeholders are those who could benefit the most:

• Researchers wanting to share, get credit, and reuse each other’s data and interpretations

• Professional data publishers offering their services

• Software and tool-builders providing data analysis and processing services such as reusable workflows

• Funding agencies (private and public) increasingly concerned with long-term data stewardship

• Data science community mining, integrating and analysing new and existing data to advance discovery. 

Improving the quality, accessibility and management of data sources across the industry will become a valuable resource for many organisations. Not only could FAIR data be used to address the challenges of long development timelines, but also prevent problems arising in the future, thanks to the shared knowledge of others. 

Drug development relies so heavily on participant information and the findings from previous trials which emphasises the importance of re-usable data to significantly cut down the cost and time of R&D, ultimately shortening the time to market for new and improved drugs. 

The concept of reusing and improving accessibility to shared data would also allow pharma organisations to respond more rapidly and effectively to changes in business demands – in comparison to information silos and generating new data which is time consuming and costly. 

Example of FAIRness 

Open PHACTS is a data integration platform for information pertaining to drug discovery. Using a machine-accessible interface, users can access the platform which provides human and machine-readable representations.  

One of many uses for Open PHACTS is obtaining gene names correlated with UniProt identifiers. UniProt is a freely accessible, online database of protein sequences and functional information. The related proteins retrieved from these methods may represent splice variants, orthologues or homologous paralogues.

The platform draws together multiple sources of publicly-available biomolecular, pharmacological and physicochemical data which can respond to “structured, well defined queries in a meaningful and reproducible way”. 

Charlotte Di Salvo, Lead Medical Writer
PharmaFeatures

For more articles covering the pharmaceutical industry, clinical research and academia, visit our content site PharmaFeatures.