Although SAS is the main player in the pharmaceutical industry, R is gaining popularity. In this post I will just mention some initiatives, documents and podcast episodes dealing with using R in pharmaceutical research and development programs.

Recently, it was a pleasure to listen the episode of the Effective Statistician podcast about the rise of R and the role played in it by the Application and Implementation of Methodologies in Statistics (AIMS) group within the asociation of
Statisticians in the Pharmaceutical Industry (PSI).

I think that it is worth to mention once again that FDA has clearly affirmed in the Statistical Software Clarifying Statement (6 May 2015) that it is not required the “use of any specific software for statistical analyses, and statistical”. However, the software used for statistical analyses “should be fully documented in the submission, including version and build identification”.

The main objectives of the AIMS group about R in pharma are educating statisticians and statistical programmers and validating and documenting R. The R validation project documentation is stored on a dedicated GitHub repository.

The second annual R/Pharma conference will take place this year in August at Harvard University. The call for papers give an idea of the opportunities and challenges of the use of R in pharma: reproducible research, regulatory compliance and validation, safety monitoring, clinical trials, drug discovery, research & development, PK/PD/pharmacometrics, genomics, diagnostics, immunogenicity.

A useful resource, dense of suggestions and ideas, is a document issued by Adrian Olszewski.

Cytel provided in a blog post a list of clear pros of R compared to SAS: ability to create effective visualisations; flexibility to combine with other tools; quick release of cutting edge methods; naturality of being a suport for collaboration and therefore innovation, as open source tool.

In conclusion, the rise of R in pharma is a fact and a lot of initiatives are in place. However, this rise is slow and mostly limited to visualization and machine learning applications, due to the position gained by SAS as validated standard in the industry.