Background
Software is becoming more complex every day, and the need to deliver fast is pushing software organizations and teams to take sub-optimal decisions instead of prioritizing solid software, creating technical debt.
We know that technical debt is partially managed in software industry and can generate an average waste of around 30-40 % of development time [1]. But what is the situation of specific fields such the bioinformatics one where software has a different life cycle, development process and stakeholders?
Scientific software is most often developed by scientists without formal education in software engineering, and its role is rarely recognized in delivering valid scientific findings [2]. Although bioinformatics software guides medical decisions, helps us identify new drug targets, and characterize diseases, most used packages struggle to mature as a product [3].
Shedding light on technical debt status in bioinformatics can help the field grow more awareness about its software products and the current challenges to overcome.
[1] Terese Besker, Antonio Martini, Jan Bosch, “Software developer productivity loss due to technical debt—A replication and extension study examining developers’ development work” Journal of Systems and Software, 2019
[2] Carole Goble, "Better Software, Better Research," in IEEE Internet Computing, 2014
[3] David Matthews “Ex-Google chief’s venture aims to save neglected science software” Nature Technology Features, 2022
Aim
This thesis aims to understand what are the challenges and the consequences related to accumulating technical debt when developing bioinformatics software and tools.
Method
This thesis includes exploratory work using qualitative methods such as interviews or quantitative such as questionnaires. Output from static analysis tools can also be used to corroborate results. The outcome will be a taxonomy of the most frequent and impactful causes and consequences of technical debt occurrences in the bioinformatics field.
Collaboration and supervision
The thesis is in collaboration between software engineering and bioinformatics (especially The Mathelier Group) and the student will be co-supervised by one member from each group.