The statement
“Were Pfizer to use R in clinical studies, it would run the risk of seeing its research questioned or even rejected by regulators doubting the veracity of results based on what they view as an unknown quantity. ”
is quite misleading. As another post said, FDA places no restrictions on the software used for drug and device registration. And for those who feel (improperly, in my personal opinion) that federal regulations on software used for database engines and medical devices apply also to analytical software, CFR11 compliance for the base R system has already been done.
The increasing use of R within FDA is perhaps testimony enough.
Frank Harrell
Chairman, Biostatistics, Vanderbilt University
Chairman, Biostatistics, Vanderbilt University
-----------------------------------------------------------------------------------------------------------------------------------------------------------------
Many of the issues raised thus far with using R in commercial environments (including use of R in regulated areas such as clinical trials) are no different than with other open source software which is widely used. Such issues are generally addressed by commercial open source software vendors. In fact, the commercial open source business is relatively mature, so it’s surprising there has been no reference to this.
Pharmaceutical companies, banks and government agencies around the world are running Red Hat Linux, for example, one of the earlier examples of a commercial open source model. It’s difficult to think of something more fundamental in computing than the operating system. Whilst such companies are still using “Linux”, they are in fact using binary builds from vendors such as Red Hat who are providing a documented, supported build of Linux together with recourse in the event of issues with the software, just like any proprietary software vendor. REvolution Computing is the same in its relationship to R in all these respects.
In such cases, open source software (as opposed to simply “freeware”) is a positive advantage, in that the source code is transparent to both practitioners and regulators – not to mention subject to the oversight of a worldwide community. In the case of REvolution’s binary build of R, REvolution R, support for validation is no different (if not better) than with proprietary software, in that the software is built to validation standards in a controlled and documented software development lifecycle. Of course, we also provide support to our customers in this regard (many of whom were referenced in the original article) in addition to performance benefits when running R internally.
-----------------------------------------------------------------------------------------------------------------------------------------------------------------
I work in the field of spatial data analysis and GIS. During my PhD study time (2000-2003), I had a lot of ideas and things I wanted to experiment with (and I did have licenses of commercial software ArcGIS, Erdas Imagine etc.), but somehow, I had problems putting my idea into work. Then, in 2004/05 I started with R. At about the same time, a group of colleagues (http://cran.r-project.org/view=Spatial) developed a support for spatial data in R. Now, 70% of my work goes through R (and we do big projects). I also use it for teaching (no complains so far; as long as you write up good lecture notes and take the right steps). I am looking forward to developing a package or two in the coming years.
I think that there are three main reasons why R makes a big difference: (1) transparency – it allows me too look at the algorithm and data (object) structures; and extend them; (2) quality – many of the (spatial and environmetrics) packages are developed by leading scientists; and I do trust academia more than fancy commercials; (3) enthusiasm – people are ready to share ideas and experiences, without a special ambition to make large profits, to monopolize or to put their personal interest in front (I still do not know how to explain this).
Articles like this help us convience also our colleagues to consider using this software (switching to R). It really is a noble task!
-----------------------------------------------------------------------------------------------------------------------------------------------------------------
For economists and especially financial analysts there is a wonderful series of R packages in Rmetrics.
-----------------------------------------------------------------------------------------------------------------------------------------------------------------
SPSS since Version 16 in 2007 has embraced R. SPSS users can install a free plug-in enabling them to run R code on the active SPSS data with output appearing along with regular SPSS output in the Viewer window.
R and SPSS have complementary strengths. R packages can extend the statistical capabilities of SPSS while SPSS can handle heavy-duty data management and present the R results in a high-quality output format. Users can define SPSS-style syntax for R packages and easily build custom dialog boxes for them (as well as other things), so users who don’t want to make the not insignificant investment in learning R can still take advantage of the many available R packages. Details can be found at SPSS Developer Central
-----------------------------------------------------------------------------------------------------------------------------------------------------------------
The burgeoning interest in R demonstrates that there’s demand for analytics to solve real, business-critical problems in a broad spectrum of companies and roles, and that some of the incumbent analytics offerings, in particular SAS and SPSS, don’t sufficiently meet the growing need for analytics in many major companies.
S+ (now TIBCO Spotfire S+, http://spotfire.tibco.com/) is a commercial software package based on the S language, which was a forerunner of R as discussed above, and has been widely adopted. It is currently used in a wide variety of areas, including Life Sciences (including the analysis of Clinical Trial data), Financial Services, and Utilities, for applications such as speeding the analysis of clinical trial data, optimizing portfolios, and assessing potential sites for building wind farms.
We welcome, respect, and appreciate the vitality, creativity, and sheer productivity of the R community, and the high quality of statistical methods the community creates. Because of the close historical ties between the two products, it is generally easy to port most R statistics into the commercial S+ environment, and we have worked to make that easier in recent releases.
Once in S+, these analytic methods can be incorporated into intuitive tools for business decision makers and deployed to automated environments, using visual workflows, web-based applications (using standard web services), Spotfire Guided Applications for dynamic visual analysis, and scalable, event-driven architectures using TIBCO’s IT infrastructure. S+ also provides some unique offerings, such as the ability to flexibly and efficiently analyze very large data sets. In this way, we feel companies can maximize the value of their analytic investments to make rapid business decisions, whether those analytics are developed in R or S+.
