De novo protein design is the holy grail of synthetic biology. The combination of more powerful computation and the decreased cost of
synthesizing DNA has allowed in silico protein engineering to provide a path to customizable proteins. The method explored here does
Bayesian optimization using a Gaussian process with an edit distance kernel to allow for informed exploration of novel sequences starting from known protein sequences. After training the model on diverse datasets that captures both structure and function, we will use it to design a novel mu opioid receptor. Natural mu opioid receptors bind to any opioid related molecule; an opioid receptor will be engineered to selectively bind to fentanyl, a potent and lethal opioid. This protein can then be incorporated into a microfluidic sensor to inform first responders about aerosolized fentanyl.

I have been working in the Weiss lab since my freshman year as part of the MIT iGEM team. My experience there instilled in me the
love of research and a passion for synthetic biology. I look forward to the opportunity SuperUROP will give me to do original research and prepare for graduate school. This project has the potential to save lives, and I feel honored to take part in such important work.