Hugh Y
- Research Program Mentor

MD/PhD candidate at University of Chicago

Expertise

Machine learning for clinical applications, Protein engineering, Immunoengineering, Cancer immunotherapy, Medicine, Biology, Computer Science

Bio

I am broadly interested in tool-development at the intersection of machine learning, medicine, and immunology. I previously worked in the area of synthetic biology with an emphasis on genetic engineering via CRISPR-Cas9 systems and protein engineering. My more recent projects include designing machine learning models to predict patient outcomes, applying machine learning frameworks to study inequities in healthcare, and building models to aid in protein engineering. In general, I am interested in designing impactful tools that can improve our understanding of human health and push medicine forward. Outside of my research, I consider myself an avid outdoorsman. I love to hike, bike, and canoe/kayak. My most recent trips include hiking trips to the Grand Canyon, Macchu Picchu/Huayna Picchu, and Rainbow Mountain. Additionally, while I am definitely not an expert, I am very interested urban artwork.

Project ideas

Project ideas are meant to help inspire student thinking about their own project. Students are in the driver seat of their research and are free to use any or none of the ideas shared by their mentors.

Predicting protein-based drugs efficacy on treating lung/breast cancer.

In recent years, there has been an explosion in the development of protein-based drugs. Many of these drugs have been designed to target cancer cells by inhibiting proteins necessary for cancer cells to proliferate. While current protein-based drugs for cancer has significantly improved patient's prognosis, these drugs remain challenging to design. Over the last few years, publicly available data on drug efficacy on several different targets has been accumulating. Using this data, one could analyze patterns that make certain drugs more effective than other drugs and even predict drug efficacy based on structural and chemical properties of these drugs.

Predicting co-viral infection in SARS-CoV-2 patients and the outcomes of coinfection

While we now have relatively fast diagnostic tests to determine if a patient has COVID-19 or not, patients testing positive may also have another viral infection that may have similar symptoms to COVID-19 and thus may not be detected. An example of this is a patient with both a COVID-19 and a influenza infection. Such coinfections may lead to worst symptoms outcomes and patients may need additional care. As such, it would be useful to have a model to predict if a patient is likely to have a coinfection (and if so, what virus?), and to study the impact of such co-infection. This would allow physicians to allocate different levels and types of resources in order to improve outcomes and the severity of symptoms.

Curating recent advancements of CAR-T therapy.

Chimeric antigen receptor (CAR) T cell therapies has significantly improved outcomes in cancer patients. Over the last several years, there has been many variations of CAR-T therapy, such as armored CAR-T therapy, dual CAR-T therapy, self-destruct CAR-T therapy, etc. Additionally, CAR-T therapy has been applied to other diseases. It would be very impactful to curate these different variations, the pros/cons of each variant, and the use case of each of these variants.

Coding skills

python, R

Teaching experience

My teaching experience began during my time as an undergraduate where I worked as a tutor for student-athletes and taught core mathematics, biology, chemistry, and physics classes. During this time I also volunteered at underserved elementary schools and high schools as a teaching assistant. Additionally, while working at Microsoft, I volunteered at outreach programs to conduct demos and help teach students basic coding. Currently, I am helping lead a group of medical students through a project focused on applying machine learning to predict cancer patients risk scores.