Introducing the FAIR Principles for research software

The FAIR Guiding Principles for scientific data management and stewardship, published in 2016, aimed to make valuable scientific outputs more Findable, Accessible, Interoperable, and Reusable (FAIR). Recognizing the crucial role of research software, the FAIR4RS Working Group—a collaboration between the Research Software Alliance (ReSA), FORCE11, and the Research Data Alliance (RDA)—adapted these principles specifically for research software. This adaptation acknowledges the unique characteristics of software, such as executability, composite nature, and continuous evolution, necessitating revisions to the original FAIR principles. The FAIR4RS Principles are intended to benefit all research stakeholders by promoting transparency, reproducibility, and reusability. This paper emphasizes the significance of the FAIR4RS Principles and showcases their positive reception within the research community. While addressing research software and data discoverability is an ongoing challenge with prior efforts, FAIR4RS provides a unifying framework integrating past initiatives.

The paper references previous work on research software discovery (Struck, 2018), research objects (Bechhofer et al., 2010), and efforts to apply FAIR principles to research software (Katz et al., 2021). It highlights the need for a unified approach that considers the unique aspects of research software and its development ecosystem (social coding platforms, package management systems). The authors note that unlike data, which is often FAIRified upon publication, research software ideally should strive for FAIRness from its initial development stages, given its potential for direct use.

The FAIR4RS Principles were developed through extensive community consultation, involving approximately 500 individuals from over 110 organizations and 34 countries. Subgroups worked from July 2020 to March 2021, producing outputs that informed the principles' development. A draft was released in June 2021 for community review. The final version, summarized in the paper, represents the culmination of this collaborative effort. The process leveraged and amplified existing community momentum by integrating various past efforts to apply FAIR principles to research software. The FAIR4RS Working Group's success stemmed from bringing together diverse initiatives and aligning with existing FAIR data efforts, thus building upon pre-existing community engagement and knowledge.

The paper presents the FAIR4RS Principles (version 1.0), organized under the four core FAIR principles: Findable, Accessible, Interoperable, and Reusable. Each core principle is further detailed with numbered guiding principles. The definition of research software provided by FAIR4RS is given. Three examples illustrate the principles' practical application: Comet (a command-line tool for tandem mass spectrometry), PuReGoMe (a collection of Python scripts and Jupyter notebooks for analyzing Twitter data), and gammaShiny (an application providing graphical user interfaces for the R gamma package). These examples showcase how globally unique and persistent identifiers, rich metadata, standardized protocols, open licenses, and detailed provenance contribute to FAIR research software. The paper highlights successful initial adoption by organizations such as ELIXIR, the Australian Research Data Commons (ARDC), the Netherlands eScience Center (NLeSC), and ZB MED. These organizations are implementing the principles through various means, including policy changes, development of guidance materials and self-assessment tools, and incorporation into software management plans.

The development of the FAIR4RS Principles signifies a crucial step in the research community's recognition of research software's inherent value. While the FAIR Guiding Principles can be applied to research software by treating it as a digital research object, the FAIR4RS Principles offer a more nuanced and tailored approach that directly addresses the unique challenges and characteristics of research software. The community-endorsed nature of the principles fosters shared practices and increased awareness of the challenges faced by various stakeholders. The paper highlights the positive impact of the FAIR4RS Principles, ranging from increased clarity for funders to improved guidelines for publishers and institutions on software sharing requirements. The successful initial adoption by key organizations demonstrates the principles' relevance and value.

The FAIR4RS Principles provide a robust framework for enhancing the Findability, Accessibility, Interoperability, and Reusability of research software. The wide-ranging community consultation and successful initial adoption demonstrate the principles' significance and relevance. Future work should focus on increased standardization of community practices in areas such as metadata, identifiers, and domain-relevant community standards. The FAIR4RS Principles serve as a vital tool for promoting transparency, reproducibility, and sustainability within the research software ecosystem.

The paper acknowledges that achieving widespread adoption of the FAIR4RS Principles is an ongoing process. The success of the principles ultimately depends on the commitment and engagement of the research community, along with the availability of supporting infrastructure and tools. While the paper highlights several successful implementations, further evaluation and monitoring are needed to fully assess the long-term impact of the FAIR4RS Principles.