While being able to reproduce results is a major achievement, transparency is equally important: the validity of results can only be fully assessed if the parameters, software and custom code of each analysis step is fully accessible. On the level of the code, a data analysis therefore has to be readable and well documented. On the level of the results it has to be possible to trace parameters, software stack and code through all involved steps.
Finally, valid results yielded from a reproducible data analysis become even more beneficial for the scientific community once the analysis can be reused for other projects. In practice, this will almost never be a plain reuse, and instead requires adaptability to new circumstances, e.g. being able to extend the analysis, replace or modify steps and adjust parameter choices. Such adaptability can only be achieved if the data analysis can easily be executed at a different computational environment (e.g. a different institute), thus it has to be scalable and portable again. In addition, it is crucial that the analysis code is as readable as possible such that it can be easily modified.
In this work, we show how sustainability in terms of these aspects is supported by Snakemake. Since its original publication in 2012, Snakemake has gained a wide adoption, culminating in, on average, nowadays more than 3 new citations per week, and over 600 citations in total (Fig. \ref{662385}), making it one of the most widely used workflow management systems in science.