Hi all! My name is Mary Lavin and I worked in the Doyle Lab at UCLA as a Data SURF student through C-CAS this summer. My project focused on building machine learning models for structure-property relationship modeling of organic photocatalysts. Leading up to this internship, I was nervous about my lack of machine learning experience and felt uncertain about how much I would be able to accomplish. Reflecting on the past couple months, I’m very proud of how much I have grown as a computational chemist. I created an end-to-end machine learning pipeline for important photocatalyst properties such as absorption/emission (nm), ground state redox potential (eV), and excited state redox potential (V), all with mean absolute errors comparable to DFT calculations. I also synthesized novel photocatalysts for experimental validation.
This experience has helped me further my wet lab skills and learn new computational techniques. Furthermore, I received invaluable advice from the graduate students in the Doyle Lab, which will continue to inform how I approach my graduate studies. Outside of work, I enjoyed the new experience of living in LA, especially as someone who grew up around the East coast. I went on hikes, explored the city, and spent time at the beach. Thank you to Professor Abigail Doyle, my mentor Jason Wang, Courtney Holland, and C-CAS for this experience! I had a great summer and I’m excited to see what C-CAS continues to accomplish.
--- Mary Lavin, 2023 participant