Building a Rapid Drug Discovery Pipeline Specific for Pediatric Acute Myelogenous Leukemia with Machine Learning/In Vitro Feedback Loops
Abstract
Fiscal year 2022 (FY22) Peer Reviewed Cancer Research Program (PRCRP) Topic Areas: This research proposal addresses the two Topic Areas of blood cancer and pediatric, adolescent, and young adult cancers. Scientific Objective: The overarching objective of this proposal is to use artificial intelligence to develop new and safer drugs to treat a deadly type of childhood blood cancer: acute myeloid leukemia (AML). Scientific Rationale: The need for safe and effective ways to treat children with AML. Pediatric AML is a deadly form of cancer and is responsible for nearly half of all deaths among children with leukemia. Indeed, 4 out of every 10 children with AML die within 5 years of being diagnosed with this devastating disease. Chemotherapy for children with AML involves extremely intense and very toxic treatments that cause heart and fertility problems and can even cause other forms of cancer later in life. Since currently available treatments for AML are so toxic, doctors have little ability to adjust therapies to make them more aggressive to treat the most high-risk patients. This illustrates a critical need for safer and more effective options to extend and improve the lives of children with AML. Eliminating cancer-causing fusion oncoproteins to treat AML: Our team predicts that the key to creating safer and more effective treatments for AML lies in the ability to eliminate fusion oncoproteins from the body. Fusion oncoproteins are a unique type of protein only found in cancer cells that form when portions of two otherwise separate proteins combine (fuse) to form a new, cancer-causing protein. One of the key reasons that children develop AML is because of these fusion oncoproteins; therefore, fusion oncoproteins are ideal drug targets because they transform healthy cells into cancerous cells. However, fusion oncoproteins are nearly impossible to eliminate with drugs using current drug-development approaches because of their unique shape and activity. Thus, new approaches are needed for scientists to develop drugs that can bind to and eliminate fusion oncoproteins. Our plan to develop new drugs that target and eliminate fusion oncoproteins: To develop new drugs capable of eliminating fusion oncoproteins from the body, there are three important steps that need to be overcome that we will address in our proposed project. Firstly, we need to determine the shape of each fusion oncoprotein (Aim 1); secondly, we need to identify compounds that can bind to the unique shape of each fusion oncoprotein (Aim 2); and lastly, we need to test how well these identified compounds can bind to and eliminate fusion oncoproteins from cancer cells (Aim 3). Since traditional approaches to determine protein shape and identify binding partners are highly inefficient, we will harness the power of artificial intelligence to produce a new algorithm that can predict the shape and potential binding partners of fusion oncoproteins. We will then test how effectively the compounds that we identified using our algorithm can bind to and eliminate fusion oncoproteins from special leukemia cell models developed by our team. FY22 PRCRP Overarching Challenges: Our proposed project addresses the overarching challenge of Therapeutics. The current drug development process in AML involves a trial-and-error approach that tests one potential drug at a time; as such, this process is slow, inefficient, and has left drug discovery in AML stagnant for nearly 40 years. We propose that the development of a new and improved drug discovery pipeline, which combines the use of artificial intelligence to first predict novel drug candidates with experimental cell models to then test these candidates, will lead to the efficient and rapid design of safe and effective drugs that eliminate fusion oncoproteins from cancer cells. Ultimately, our work has the potential to transform AML treatment options by (1) developing the drugs able to eliminate fusi
Document Details
- Document Type
- DoD Grant Award
- Publication Date
- Jan 04, 2024
- Source ID
- HT94252310706
Entities
People
- Kolja Eppert
Organizations
- McGill University Health Centre
- United States Army