Project title: “Deep Mutational Scanning of Receptor Tyrosine Kinases Involved in Cancer Using Transfer Learning Approaches”

Research team includes: Erik Procko, assistant professor in biochemistry

What led to your exploration of cancer? 

Cancer is one of the major challenges facing humanity. I started working on cancer-related problems during post-doctoral research at Stanford University mainly due to the potential impact that my research could have on our community.

Can you provide a short summary of your project? 

The main goal of the project is to understand how human growth factor receptors, which play a critical role in numerous cancers could be selectively targeted using drugs.

How will your project advance cancer research?

Our goal is to demonstrate how cutting-edge experimental techniques and machine learning approaches can provide information that is inaccessible to current experimental approaches for a key class of cancer-related proteins. This approach is generalizable to other cancer-related targets to provide key information for understanding and tuning the selectivity of drugs.

How did the CCIL Seed Grant benefit your research?

It gave my lab an opportunity to expand our computational research into a new area related to growth factor receptors, which play a critical role in cancer.

Was there anything you were able to accomplish, that you might not have been able to, if you had not received the grant?

The grant allowed us to develop the collaboration with Erik Procko’s lab, which is pioneering the use of deep-mutational scans for understanding the fundamental mechanisms of disease regulation and prevention.

Why is it important to support cancer research?

Drug design is one aspect of cancer research that gets maximum attention. However, understanding the causes of cancer is a huge area of research and has wide implications for cancer prevention in healthy individuals.

What areas of cancer research/technology do you hope to explore more as a result of your research findings?

Integration of data-science based approaches for the design of cancer drugs is one of the main areas for future research. Through this project, we have shown how transfer learning can be used synergistically with high-throughput experiments to tackle some of the big challenges in this field.

Are there any plans for clinical or industry translations of these projects in the near future? If so, what do they look like? 

We don’t have any plans in near future. However, my co-PI, Erik Procko, has used the deep-mutational scans of the growth factor receptors to engineer a highly effective treatment for the cancer-associated virus human cytomegalovirus (HCMV), which can be potentially commercialized.