Portal de Periódicos da CAPES

There ain't no such thing as a free lunch – especially in physiology software development

2023; Wiley; Volume: 601; Issue: 21 Linguagem: Inglês

10.1113/jp285573

1469-7793

Christopher G. Wilson,

Advanced Chemical Sensor Technologies

Physiologists, like many scientists, have a long history of developing DIY apparatus for their research. A canonical example of this was published in this journal in 1952. Hodgkin, Huxley and Katz detailed the recording system they developed to record action potentials from squid giant axon (Hodgkin et al., 1952). The tools they developed led to later, Nobel-prize winning, work on understanding the action potential. The 21st-century version of this kind of tool development includes the production of bespoke software for a laboratory. In this issue, Lusk et al. (2023) describe their whole-body plethysmography software that dramatically increases the efficiency and throughput of data analysis. Breathe Easy provides easy import of data from data acquisition software, automated breath detection, comprehensive data analysis, and output of publication quality graphics. Breathe Easy provides a sophisticated automated respiratory data pipeline for whole-body plethysmography waveform detection and analysis. A very important feature of Breathe Easy is that it is free open-source software (FOSS) – in this case, based on Python and R – that incorporates current best practices such as parallelized data handling, modularity, and a foundation for portability (it currently runs on Windows but can be compiled to run on MacOS and Linux as long as dependencies are met). The authors provide a comprehensive overview of their high-throughput, open-source software, and examples of data analysis using the suite of modules in Breathe Easy. Their work is in mice but the tools look usable across a wide range of mammalian models. The authors’ software has the potential to vastly speed up data tagging and analysis for plethysmography. Breathe Easy represents a significant software development project that dramatically benefits investigators using plethysmography in cardio-pulmonary physiology research. Thus, Breathe Easy provides an inexpensive and powerful addition to the acquisition hardware/software armamentarium for physiologists. Much of the scientific software developed over the last two decades falls under the category of FOSS. Using FOSS encourages reproducibility and, arguably, the majority of scientific software should be open-source (Fortunato & Galassi, 2021). FOSS is distributed with the idea that the user is not required to pay for the software, has access to the source code, can modify the software if they choose, and can distribute it without cost. However, in this context, ‘free’ means unencumbered by restrictive licensing, rather than ‘without cost’. Thus the saying, ‘there ain't no such thing as a free lunch’, or TANSTAAFL – in use for a century but popularized in Robert Heinlein's The Moon is a Harsh Mistress (Heinlein, 1997) – still applies to FOSS. The cost of FOSS is the effort needed to learn the software and fit it to a laboratory's needs. FOSS often lacks comprehensive documentation or the support received from proprietary vendors. So it is rare to find FOSS scientific software that provides broad usefulness and sophistication as well as high quality documentation, like Breathe Easy. The success of ImageJ, originally written by Wayne Rasband at the National Institutes of Health (USA) sets a high bar for open-source scientific software. ImageJ is extensible via macros and plug-ins, and was designed this way to ensure customisability. Allowing users to modify software and have access to the source code makes it possible to rapidly expand the usability and stability of scientific software (Schneider et al., 2012). Because ImageJ is FOSS and widely used, it was adopted as the foundation for the Open Microscopy Environment (http://www.openmicroscopy.org) and has spawned successor products such as Fiji, Micromanager and ImageJ2. ImageJ is the epitome of excellent open scientific software. Similarly, Breathe Easy’s architecture is open-source (Python and R), embodies modular programming practices, making it possible for the community to expand upon the work that Lusk and colleagues have already invested in Breathe Easy. In contrast to the success of ImageJ and its offspring, some FOSS projects should be seen as a cautionary tale: the plethora of ‘abandonware’ strewn across the Internet represents many very useful software development efforts that have become fallow and are no longer supported. There are two major forces at work that result in abandonware: (1) an individual in a laboratory has a talent and thirst for developing their own tool but the tool itself is focused on a narrow range of applications associated with that laboratory's work so it does not create a thriving user base; (2) very few physiology/biology laboratories have the resources to sustain high-level software development and marketing budget or social media presence to attract new user-developers. To revive these projects would require an intra-laboratory champion and the resources to support that person's work on the software as well as an engaged community to expand and improve the software. There is hope, however, for physiologists who wish to use FOSS tools. Breathe Easy is an excellent example of well-engineered software that is useful across a wide range of laboratories focused on cardio-pulmonary physiology. Dr Lusk and colleagues have produced software that uses contemporary development methods, software architecture, solicits feedback from users, and is under active development. It works now and, with sufficient support from users and funding bodies, can serve as an extensible analysis platform. Regardless of what lies ahead for Breathe Easy, the availability of high-quality FOSS that automates data analysis provides a significant time and efficiency force-multiplier for experimentalists. Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article. None declared. Sole author. NIH R01 AT011691-01 Unraveling Lung Interoception and its Functional Consequence in the Developing Ovine Lung.