While the open source statistical language R still has a way to go to gain the market cachet of its predecessor, S, one startup software firm is working to change that.
Random Technologies, an offshoot of University of Rochester Medical Center, launched its new statistical-analysis software package, coined RPro, at the recent Drug Information Association conference in Atlanta. RPro is designed to support a number of R-based analysis tools of interest to life science researchers, including the open source microarray analysis package Bioconductor.
Yet RPro, while perhaps designed to compete with S-Plus, the commercial implementation of S marketed by Insightful, has a different modus operandi.
Greg Warnes, an associate professor of biostatistics and computational biology at the University of Rochester and an active R developer, noted that RPro has a particular edge over S-Plus in the case of Bioconductor. “One of the advantages we have over Insightful is because we are actually using R as opposed to S, we can keep much closer to the release cycle of the Bioconductor developers,” he said.
“Insightful, because they must take what Bioconductor develops and [translate it] into [the] S language, building all the graphical language tools … [results in] a substantial delay,” Warnes told BioInform.
Duncan West, product manager at Insightful, indicated that the firm doesn’t view Random Technologies as a competitor, but rather, as proof that statistical software is gaining importance with the life science research community. “We hope they fare as well as we have,” he said. “Any gains for the S languages in biopharm are a good thing.”
West noted that Insightful has made improvements in S-Plus that “allow for the use of both S languages with the Insightful product. We have been working with some of the core Bioconductor developers and other developers of R to make porting into S-Plus easier.”
Indeed, the latest version of Insightful’s software, S-Plus 8, includes a feature called the S-Plus Package System that allows users to convert R packages for use within S-Plus to enable “cutting-edge statistics … written in open source” to work “within a commercially supported environment.”
Random Technologies has a similar goal, and RPro was designed to provide commercial-grade support for what is considered one of the most widely used statistical computing and graphics systems in biomedical research. The company said it plans to standardize its process for releasing, upgrading, and adding new features to RPro, and will offer technology support, documentation, and validation for commercial users, particularly in the area of regulatory compliance.
First Red Hat, Now Random?
Warnes said in a company statement that Random intends “to do for R what Red Hat did for Linux.”
In conversation later with BioInform, Warnes said that Random added several features to the core R software system that are of interest to people who need commercial support “but are not particularly concerned about the open source community.”
“One of the advantages we have over Insightful is that since we are actually using R as opposed to S, we can keep much closer to the release cycle of the Bioconductor developers … Insightful, because they must take what Bioconductor develops and [translate it] into [the] S language, building all the graphical language tools … [results in] a substantial delay.”
One such feature is the relatively easy way an individual can take an existing functionality — for example, a package with test scripts and example data — and tailor it for distribution via RPro. Warnes noted that there are approximately 1,200 add-on packages for R available through the Comprehensive R Archive Network. The current version of Bioconductor, meanwhile, includes more than 200 software packages, more than 1,100 metadata packages, and 44 packages of experimental data.
“There is a demand for an organization [to be able to use] tools [that] meet certain requirements. We are [tapping into] the overlap between the demand from users and [the needs of] an organization,” Warnes explained. “The value of the tool, RPro[is designed] … to deploy statistical computations where they previously have not been gained,” he said.
Random does face one challenge, however, in that open source remains relatively unproven in terms of an ancillary commercial product market. Warnes said that for 20 years, though, Insightful’s products have made inroads in the commercial marketplace, indicating that this is not at least a major concern for his company.
When asked how he envisions the commercial S-Plus faring against the open source R, West said “I can’t speak for the R community, [but] we’ve had a good response as we’ve worked with package authors and oftentimes at suggesting changes that could be made so their packages can work in both R and S-Plus.”