Computational Approaches for Cancer – Looking Around the Corner
Eric Stahlberg, PhD
Director, Biomedical Informatics and Data Science, Frederick National Laboratory for Cancer Research
Looking ahead to the sixth Computational Approaches for Cancer Workshop (CAFCW) at SC20, I am extremely excited about how the HPC community has embraced cancer research and how far we’ve progressed in such a short time. Five years ago, when the first CAFCW was being planned, the U.S. national initiative for precision medicine was less than 6 months old; the National Strategic Computing Initiative had not yet been signed; exascale systems were still in development; and the collaboration between two world leaders in computing and cancer research, the U.S. Department of Energy (DOE) and the National Cancer Institute, had yet to be established.
Today, due to the immediate benefit to researchers, clinicians and patients alike, high performance computing is now driving innovation across cancer research and clinical care. Previously unimaginable exascale technology systems are coming online and are being used increasingly for cancer research. Collaborative team science is becoming the new norm—motivated by biomedical science and forward-looking public policy across the globe. This includes support for computational biomedicine across the European Union, Asia and North America; access to increasing amounts of open data; and U.S. public policy such as the Precision Medicine Initiative, the National Strategic Computing Initiative, the 21st Century Cures Act and related Cancer Moonshot℠ efforts and the DOE Exascale Initiative.
Most recently, the phenomenal growth in the use of artificial intelligence across biomedicine is transforming and revolutionizing the role of high performance computing in cancer research and clinical care.
While HPC has enabled biomedical research for years, the work highlighted at last year’s SC19 conference in Denver was noteworthy in several ways. Across SC19, cancer and biomedical computing workshops, presentations, birds of a feather sessions and panels infused the program and captivated new audiences across all generations.
The 5th Computational Approaches for Cancer Workshop was again filled, with the largest number of submissions ever shaping the workshop. A breakthrough multidisciplinary team of DOE and NCI-supported scientists at the national laboratories received the SC19 Best Paper award! And leaders from across the cancer and computing research communities attended, eager to share how the collaborative work by biomedical and HPC scientists made new insights possible. It was clear not only that HPC is vital to cancer research, but that SC is “more than HPC.”
As innovative computational approaches for cancer continue to grow and expand, exciting new frontiers come into focus. Insights from data are being transformed to become predictions; predictions are joined with real-world outcomes; and a data-driven eco-system for learning about cancer is emerging. New levels of analytics are now being used to characterize and diagnose cancer precisely and accurately. Like other fields that already utilize digital twins, growing amounts of cancer data, interdisciplinary collaborations and computational approaches are opening the door to amazing predictive models for this complex and challenging array of diseases.
This year finds us all looking at SC20 differently, amidst the global COVID-19 pandemic and again being challenged to innovate. High performance computing is being used to search for new avenues of innovation to help reduce the impact of this pernicious disease that affects all of us across the planet. Computational solutions inspired by cancer challenges are being employed in this battle. This year’s Computational Approaches for Cancer Workshop will again highlight innovations and advances in the use of computing and AI to make progress in cancer, bring together global communities and disciplines and take a glimpse into the future of cancer research in a post-COVID-19 world.
—
Eric Stahlberg, PhD, is the director of Biomedical Informatics and Data Science (BIDS) at the Frederick National Laboratory for Cancer Research. Dr. Stahlberg is a founding co-organizer of the Computational Approaches for Cancer Research Workshop (CAFCW), held in conjunction with the SC conference since 2015. He has been instrumental in establishing the Frederick National Laboratory’s high-performance computing initiative and in assembling scientific teams across multiple, complex organizations to advance predictive oncology. In 2017, he was recognized as one of FCW‘s Federal 100.