MetaR: simple, high-level languages for data analysis with the R ecosystem

Data analysis tools have become essential to the study of biology. Tools available today were constructed with layers of technology developed over decades. Here, we explain how some of the principles used to develop this technology are sub-optimal for the construction of data analysis tools for biologists. In contrast, we applied language workbench technology (LWT) to create a data analysis language, called MetaR, tailored for biologists with no programming experience, as well as expert bioinformaticians and statisticians. A key novelty of this approach is its ability to blend user interface with scripting in such a way that beginners and experts alike can analyze data productively in the same analysis platform. While presenting MetaR, we explain how a judicious use of LWT eliminates problems that have historically contributed to data analysis bottlenecks. These results show that language design with LWT can be a compelling approach for developing intelligent data analysis tools.