Home AI and ML Using High Performance Computing to Advance Medical Research
Advance medical research

Using High Performance Computing to Advance Medical Research

by Kelsey Winick

What the COVID-19 pandemic and vaccine development process have demonstrated is that it is possible to advance medical research far more quickly than thought possible as long as the funding and the infrastructure are in place. While the funding for medical research is often dependent on research grants or funding from private or public sources, the good news is that the infrastructure costs are falling. Read on to learn how researchers at UC Davis have been able to harness the power of high performance computing to advance medical research.

High performance computing (HPC) is essential technology for today’s researchers to achieve accelerated discovery and time to results. With recent advances in hardware and software development HPC has transitioned from being an esoteric application, to becoming a fundamental part of a research university. Researchers at the University of California, Davis (UC Davis) are using Oracle HPC to advance medical research that may benefit the health and well-being of millions of people around the world.

During a recent webinar, researcher Igor Vorobyov, Ph.D. in the Vorobyov Lab and Clancy Lab at UC Davis, shared how Oracle HPC is powering and accelerating his team’s research to better predict whether a drug will harm your heart. According to Vorobyov, pharmaceutical compounds designed to cure one ailment may cause heart pattern disturbances, known as arrhythmias. Not only is this type of cardiotoxicity a serious health problem for patients, it is also an expensive problem for the pharmaceutical industry. To resolve these challenges, the research team led by Vorobyov aims to predict drug-induced arrhythmia from a drug chemical structure using a multi-scale modeling pipeline.

Watch the Webinar

One of the reasons for this method was to , “reduce cost of drug development and save human lives.” The UC Davis research team’s on-premise computers needed the power of enterprise computing to process the very large micro-scale modeling pipeline datasets. They turned to Oracle for Research, which supported Vorobyov and his team with a one-year Oracle for Research grant, providing access to Oracle Cloud, technical advising and collaboration.  Using Oracle HPC, the UC Davis team was able to not only be cost effective, but also save time when processing large amounts of data. Vorobyov shared that they used machine learning applications through GPU to have “customized images and flexible storage options, while knowing the data is secure.” He added that there is minimal “wait time in this secure environment, so we can run instantaneously and also run different components of our pipeline simultaneously. We can easily automate our simulations and we have multiple sources of help and support.”

But speed and scale are not the only virtues of this next-generation Oracle HPC, it also allows for greater flexibility with storage options, including file and block. Rajib Ghosh, Global Senior Solutions Architect at Oracle for Research, explained that “object storage is being used by the UC Davis project to store a high volume of research data securely. File system storage allows researchers to seamlessly share data across the project. Block storage is a petabyte scale price performance shared storage for computation, which was also used in this project and it also provided much higher performance when storage was important.”

With Oracle HPC and Oracle for Research, the UC Davis team was able to scale-up their pipeline and use it to investigate multi-target block, mutagenesis data, and translate data between different models using data science platforms. Research accelerated by Oracle Cloud is enabling Vorobyov and his UC Davis research team to explore data in novel ways and discover solutions to better ensure pharmaceutical safety for the health and well-being of people worldwide.

Learn more about Oracle high-performance computing here.

You may also like