In the fight against cancer, early prediction, which drastically improves prognoses, is critical. Now, new research by a team from Northwestern University – and accelerated by supercomputing resources at Argonne National Laboratory – has helped to improve a technique that can enable “extremely early” cancer risk prediction.
The research team, led by Northwestern University professor Allen Taflove, sought to develop a “low-cost, high-throughput optical microscopy technique” that could detect alterations at the macromolecular level. To do this, the researchers ran “Angora,” an open-source “microscope in a computer” simulation tool.
Angora helped the researchers advance from static partial wave spectroscopy (PWS) – which examines cell activity based on snapshots – to a new form of dynamic PWS, which allowed them to more selectively and precisely observe cells’ evolution over time. “Simply speaking,” Taflove said, “we will learn much more about what goes on inside a living cell because will literally get to watch it change.”
This is all in the pursuit of the “field effect,” which describes how cancer can be detected through rigorous analysis of tissue near cancerous areas. Vadim Backman, a professor of biomedical engineering at Northwestern and co-principal investigator on the current research, was one of the first researchers to validate the field effect. Further validation and logistical support for the field effect would allow physicians to detect cancer through much more routine swabs and tests. The research team has already validated their PWS risk quantification method in hundreds of patients across a number of different organs.
To successfully run PWS – which involves observing changes as minute as 20 nanometers across millisecond resolution – the researchers turned to the Argonne Leadership Computing Facility (ALCF). They ran Angora on ALCF’s massively parallel Mira supercomputer, which is currently ranked as the 24th most powerful supercomputer in the world by the June 2019 Top500 List. Mira, an IBM system that utilizes BlueGene/Q Power 16C 1.6GHz processors, is rated at 8.6 Linpack petaflops.
In the future, PWS techniques could be used to improve treatment options – not just risk detection. The research group is already investigating the possibility of a broad-spectrum cancer treatment using dynamic PWS.
About the research
The researchers referenced in this article recently published a paper on their work, “Multimodal interference-based imaging of nanoscale structure and macromolecular motion uncovers UV induced cellular paroxysm,” in the April 2019 edition of Nature Communications. The paper was written by Scott Gladstein, Luay M. Almassalha, Lusik Cherkezyan, John E. Chandler, Adam Eshein, Aya Eid, Di Zhang, Wenli Wu, Greta M. Bauer, Andrew D. Stephens, Simona Morochnik, Hariharan Subramanian, John F. Marko, Guillermo A. Ameer, Igal Szleifer and Vadim Backman.
The original article discussing this research can be found on Argonne National Laboratory’s website at this link.
Feature image caption: Alterations of the nanoscale structure of live HeLa cells after chemical fixation, as observed using partial wave spectroscopic (PWS) optical microscopy. Chemical fixation appears to alter the cellular nanoscale structure in addition to terminating its macromolecular remodeling. (Image courtesy of Vadim Backman, Northwestern University.)