Rafael Guerrafuentes

MIT Department: Biological Engineering 
Undergraduate Institution: University of California, Los Angeles
Faculty Mentor: Ron Weiss
Research Supervisor: Sebastian Palacios
Website: LinkedIn

2019 Research Poster



I was born and raised in San Salvador, El Salvador and immigrated to the U.S at the age of eighteen. I live in Los Angeles, California, I am a junior at the University of California Los Angeles (UCLA) studying Electrical Engineering. My research interests include transmission lines, sustainable sources of energy, and the endless applications that engineering has in the medical field. I aspire to use my talents to improve the lives of the people by making sustainable energy accessible for everyone and bring power and energy to those communities in El Salvador that up to this date have no access to electricity.

2019 Research Abstract

Using Deep Learning to Predict Behavior of Genetically Engineered Cells

Rafael Guerrafuentes1, Sebastian Palacios2 and Ron Weiss3
1Department of Electrical Engineering, University of California Los Angeles
2, 3Department of Biological Engineering, Massachusetts Institute of Technology


Synthetic biology is a field that is largely dependent on trial and error due to a lot of variations and interactions at the cellular level. To save time and resources, there is a need for more tools for modeling and prediction of biological systems in the field. Although there are mathematical models that effectively predict the behavior of some genetically engineered cells, they require domain-specific expertise and are difficult to create. To address these shortcomings, the focus of this project is to the use deep learning to offer a more systematic method to predict and model the behavior of genetically engineered cells by training a neural network with experimental data to get an accurate prediction of different variations of the same genetically engineered cells. The neural network predictions for the trained data matched the mathematical model with accuracy; we expect to extend use of the neural network to predict the behavior genetically engineered cells whose data it has not being trained on. This can greatly impact the field by offering a systematic and rapid way to predict the behavior of the genetically engineered cells.