U of Chicago & ANL to Research Cancer Using AI

The University of Chicago Medicine Comprehensive Cancer Center (UCCCC) is exploring ways to slow or stop the growth of tumors that don't respond to medication by using Artificial Intelligence (AI), Machine Learning (ML), and high performance computing capabilities located at the U.S. Department of Energy's Argonne National Laboratory (ANL).

The UCCCC will receive $6 million as part of a $15 million project that will enable advanced AI/ML approaches to mine vast datasets and uncover patterns that may lead to developing new treatments for drug resistant cancers. Funding is being provided by the "Advanced Research Projects Agency for Health" (ARPA-H) within HHS.

Cancer drug discover is a complex and resource intensive process typically taking up to 15 years and more than $2 billion dollars to take a drug from the initial discovery to FDA for approval.

The joint Argonne/UCCCC project titled "AI and Experimental Approaches for Targeting Intrinsically Disordered ProtEins in Designing Anticancer Ligands" (IDEAL) will use cutting edge technology and experimental approaches to narrow down the search to only the most promising compounds that can be translated into better treatments.

"With Argonne's expertise in AI and the University of Chicago's exceptional capabilities in cancer research, we are in a unique position to solve complex scientific challenges in cancer which is one of the most pressing healthcare problems", reports Thomas Brettin, Strategic Program Manager for Argonne's Computing Environment and Life Sciences Directorate.

Researchers will use Argonne's computing and experimental facilities such as the Aurora exascale supercomputer at the Argonne Leadership Computing Facility, and the ultrabright X-rays at Argonne's Advanced Photon Source.

These technologies will allow researchers to screen billions of possible molecules including all the drugs that are currently available in just a couple of hours and simulate thousands of complexes within days

The IDEAL team will test the new model on targets known to be relevant in ovarian cancer, the deadliest of gynecologic cancers and one that is notoriously resistant to treatment. Although the pilot project will focus on ovarian cancer, the accelerated pipeline intends to apply to any target for any cancer type and be able to revolutionize the cancer drug discovery timeline.