Could the cloud save lives? It sounds like a question only a web hosting company would ask. However, the computing model – due to its improved performance, redundancy, and affordability – has made it more possible for medical and scientific researchers to analyze huge quantities of data in the search for better treatments. 2010-2013: free hosting for science Microsoft saw the opportunity early and teamed with the National Science Foundation, allowing NSF-backed projects using the IT strategy to hit the ground running. In 2010, the tech giant offered free distributed virtual servers to the NSF for three years. Jeannette Wing, on the IT staff at the NSF, said at the time that the researchers working with NSF were “drowning in data.” Using the approach was a way to get beyond the issue of infrastructure by using a model designed for effiency, accessibiity, and power. Drawing people to the solution has been an ongoing effort for Microsoft, which has funneled $15 billion into its massive infrastructure. The obvious downside to this partnership between Microsoft and the NSF was that its own proprietary software was used by the scientists to develop the tools and process the data, making it difficult to transition to the more affordable solution when the free service concluded last year. Nonetheless, researchers were made more aware of how the system could be used to expedite sophisticated research. The cloud & implications for medicine The Association of American Medical Colleges reported in 2013 on the “promise” of this strategy for medical research. The AAMC underscored that because of the way the solution is structured, it is typically billed on-demand – so medical researchers and others are able to cut their costs by only paying for the resources they need at any given time. Direct costs are not the only way that the approach has proven economical and user-friendly for universities and research centers. It has also allowed them not to have to expand their internal IT departments or their own data centers to keep pace with the big data that today’s technology makes possible. Jeffrey C. Fox, PhD, a dean in the computer science school at Indiana University, notes that research is an excellent match for this form of computing. He particularly believes that the aspect of elasticity is compelling, offering the capability of “1000 computers to analyze your data” if needed, as needed. He also notes that because the model does not run into limitations as occurs with a particular device, processing time is typically better than when using a supercomputer (the standard big-data tool in the past). For instance, Dr. Atul Butte, a professor of pediatric medicine at Stanford University, studies genomes in his research. For years the lab he directs had been operating solely through its own data center, but it has just started its transition to the virtualized platform. The first foray into the system was for a graduate student project. The complexity of the data the student was analyzing would have made the research cost-prohibitive if not for the strategy. As Butte explains, the model enabled the student to “[p]rovision the computer, get it running, and get the project done.” Other cloud research for children’s medicine Dr. Michael Cunningham’s story was covered by NPR in 2012. He has taken advantage of the IT approach as well in his research at the Seattle Children’s Hospital. His particular interest is craniosynostosis, a condition in which the skull fuses together too early in life – resulting in a misshapen head, severe internal pressure and pain, and possibly brain damage. Cunningham and others in his field have believed that craniosynostosis was probably caused by defects in bone cell communication. Using a distributed virtual system, though, he was able to propel the medical understanding forward by analyzing extraordinarily large pools of data. Cunningham and his team were able to match patients based on how closely their cells resemble each other visually. That finding seems to suggest a reasonable path forward toward an underlying cause. In turn, treatments can become more sophisticated, and an outright cure can be sought. Similarly, pharmaceutical companies are running widescale analyses of possible drug components to determine their potential for medications. One such organization (unnamed in the piece) analyzed 21 million compounds using one application. It took under four hours and cost less than $20,000. If the company had performed that study internally, it could have taken months and cost them hundreds of thousands. Speed and cost are part of what makes this form of computing attractive to researchers, but there is another crucial argument for it as well: collaboration. Because access is simple and because speed is optimized worldwide, a group of scientists can experience real-time teamwork through this model. Additionally, according to Dr. Stephen Friend of Sage Bionetworks, the strategy makes it possible for various companies to work in tandem. Friend notes that drug firms “love the de-risking that occurs” when they each agree to submit portions of numbers to a central, virtualized database. James Staten of Forrester Research mentions Patchwork Diagnostics as an example of a company utilizing the approach to advance medicine. The company has large amounts of data about various cancer tissues stored for broad access, which makes it easier to properly diagnose. The cancer database takes a number of hours to generate a result, but it allows the doctor a mathematical degree of confidence that the tissue is associated with a certain diagnosis. The cloud, research & the planet Along the lines of collaboration, and similar to the project between the NSF and Microsoft, Amazon Web Services partnered with NASA last year to make geoscience data accessible within a distributed virtual environment. The system comprises datasets of worldwide weather figures and software to compute and analyze it, allowing amateur or professional researchers anywhere to work with the data and test hypotheses. Scientific and medical research is benefiting enormously from the rise of cloud computing. The Association of American Medical Colleges has promoted the IT strategy as cost-effective and user-friendly. In medicine, it is allowing us to get closer to disease cures, perform rapid-fire analysis of potential pharmaceutical components, and improve diagnostic testing.