For long-term sustainable software in bioinformatics

The article emphasizes the critical role of maintaining research software in the sciences, due to the challenges posed by short-term funding. Drawing on our experience, we advocate for a commitment to long-term maintenance of open-source research tools, outlining the difference between maintenance and extension of software, and classifying research software into levels based on their intended use and maintenance requirements.

We present seven practices for sustainable software maintenance, including reproducible research methods, using our own tools (dogfooding), adhering to intuitive design principles, public support forums, producing clear error messages, providing easy installation via bioconda packages, and early software release for community testing.

The piece concludes with a call to action for the bioinformatics community to institutionalize commitments to software maintenance, suggesting that funders, journals, and peer reviewers should demand and evaluate maintenance plans, similarly to how data sharing mandates are currently handled. This approach aims to improve the long-term usability and reliability of research software, thereby enhancing the field's reputation and the reproducibility of scientific research.