Random Technologies, the newest start-up company to emerge from the University of Rochester Medical Center (URMC), launched its new statistical analysis software package at an international conference of drug industry professionals this week.
Random Technologies’ RPro Statistical Software is based on the open source software system “R” – the most widely used statistical computing and graphics system in biomedical research. R was developed and is maintained by a global network of thousands of volunteer programmers and has become the preferred research tool of the scientific community because of its flexibility to address novel research problems and its capacity to produce rich graphic representations of data.
However, industry has been reluctant to adopt the software to perform critical functions, such as submitting data from drug studies for regulatory approval, because the system is not commercially supported. This is similar to the circumstances that faced Linux several years ago – the open source operating system was widely used by programmers but not embraced by the business community until the company Red Hat combined the software with professional operating support and services.
“Random Technologies intends to do for R what Red Hat did for Linux,” said Gregory Warnes, Ph.D., the company’s founder, an associate professor of Biostatistics and Computational Biology, and co-director of computational software development at the University of Rochester’s Center for Biodefense Immune Modeling. “R is an exceptionally useful system with tremendous capabilities and has been widely adopted by academic and research statisticians. Random Technologies will provide the enterprise-level support that will enable corporations to deploy this software as a critical component of their business.”
Random Technologies launched its RPro Statistic Software at the Drug Information Association conference – one of the most widely attended pharmaceutical meetings in the world – which was held this week in Atlanta.
R is well suited to meet the rapidly evolving needs of the scientific community, particularly in the field of biomedical research. The advent of “genomic-scale” technologies in genetics, proteomics, and molecular biology have dramatically increased both the potential for new discoveries and the amount of information that must be processed. Only a few years ago a “large” genetics study may have examined only a dozen genetic markers. Now technologies exist that allow scientists to track hundreds of thousands of markers per study subject. Few software systems have the ability to adapt quickly to this explosive growth in data.
One of R’s fundamental strength is its flexibility to address new research challenges such as aggregating and analyzing large volumes of data or addressing novel research challenges. Because the software code is open source, researchers and programmers can develop their own add-on tools or modules that address a unique research need without having to wait for a corporate programmer to devise a solution. These new modules can then be tested and debugged by the online community of programmers and eventually become available to all users. In fact, some 1,200 such modules exist in the R system.
“One of the reasons driving the widespread adoption of R is that the software allows a person to build a tool from scratch to solve a problem that has never been solved before,” said Warnes. “This flexibility is in contrast to existing proprietary commercial packages which tend to offer a large, expensive, and rigid product with slow development cycles that cannot keep pace with the rapidly changing needs of the scientific community.”
While R’s rapid evolution is a boon to researchers, it poses a challenge for corporate information technology managers who must ensure that business critical software meets certain quality standards. Random Technologies will standardize the process of releasing, upgrading, and adding new features to its RPro software system. The company will also provide the professional technology support, documentation, and validation at a level required by both corporate users and regulatory bodies such as the Food and Drug Administration.
“Our objective is to take the best of both worlds and bring them together,” said Warnes. “The R community has done a remarkable job of creating a world-class statistical analysis and visualization tool. Random Technologies will provide enterprise grade services and support that will enable companies to harness the potential and productivity of R with the confidence provided by a reliable commercial vendor.”
In fact, there is strong evidence that the multi-billion-dollar market for statistic analysis is already moving in the direction of adopting R more widely. A memo from the pharmaceutical giant Pfizer acknowledges that R is already a significant component of its statistical core technology and that the company is in need of a supported version of the software.
Random Technologies is uniquely positioned to meet this growing demand. Warnes is well known within the R community, has been involved in the development of the software since 1997, and is one of a team of individuals who maintain the software code. Prior to joining the faculty at the University of Rochester, he worked for Pfizer as a computational statistician in the company’s Global Research and Development division and developed an insider’s perspective of the statistical computing needs of a major pharmaceutical company. During his tenure at URMC, Warnes developed much of the material that forms the basis for Random Technologies’ RPro software.