Reproducible and sharable analysis workflows

Adam Black / Research Analyst, Center for Outcomes Research and Evaluation

Maine Medical Center Research Institute (MMCRI) currently spans the spectrum of biomedical research, with basic research programs in cardiovascular biology and stem cell biology, a growing clinical and translational research program, health services and patient-centered outcomes research, and psychiatric research.

MMCRI’s mission is to enhance the health of Maine’s population through excellence in research across the continuum of the biomedical and health sciences.

"Using shared infrastructure allows users to focus on data analysis instead of on software installation and management.”

- Adam Black, Maine Medical Center

The Challenge

R is an important tool for analysts engaged in research projects at MMCRI because of the large collection of open source packages available for the language. However, constraints around installing and updating open source software packages on desktop computers in a secure environment make R difficult to use. In addition, differences in R versions and package versions between analysts make it difficult for analysts to share and run each other’s R code.

The Solution

RStudio Server Pro (RSP) allows us to create a centralized, managed R environment that all analysts can share. This method provides increased security, decreased burden for the analyst, and much less variability in system dependencies. Using shared infrastructure allows users to focus on data analysis instead of on software installation and management. The project sharing feature of RSP also allows analysts to read and run each other’s code on identical infrastructure contributing to code review and reproducibility. Collaboration becomes much easier with project sharing. The Packrat package allows analysts to further isolate analysis dependencies for specific projects, contributing to computational reproducibility.

Why RStudio?

We used RStudio’s free software for a long time before purchasing their enterprise solutions. RStudio’s investment in professionally developed free open source software benefits the analysis community at large and makes R an accessible tool for any single person or organization regardless of means. This mission along with the excellent quality of their products and support is why we use RStudio.

The Payoff

RStudio has outlined the role of the “R-Admin” and a roadmap to “bring R through the front door” of our organizations. Following this roadmap to engage with IT and set up well-supported R infrastructure in our organization has made R much easier to use. It has increased computational reproducibility and ease of collaboration for our analysts.

RStudio provides open source and enterprise-ready professional software for data science teams using R and Python.