Pierre Boyeau

I am a fourth-year PhD candidate in the EECS department at UC Berkeley, coadvised by Profs. Nir Yosef and Michael Jordan. My primary interests are in single-cell genomics, specifically in the application of deep learning and probabilistic models for single-cell data analysis. I develop statistical methods for exploring multi-modal single-cell data, and focus on delivering statistical assurances and comprehensible outcomes from deep generative models.

I have developed statistical frameworks to address multiple challenges in single cells, all of which are available as open-source software. One of these challenges is the identification of functional relationships between genes and cellular properties, which is a key step to understand the molecular mechanisms driving cell function. I have also built a deep generative model to reveal sample-level heterogeneity in cohort-scale single-cell data, critical to understand how sample variability affects cell identity. I have also developed a procedure to identify differentially expressed genes using deep generative models, which is robust to nonlinear nuisance factors.

Before my time at Berkeley, I interned in several companies to apply machine learning to real-world problems in finance and retail. In 2018, I also visited Prof. Ichiro Takeuchi at the Riken Institute in Japan, where I worked allergenic protein classification and epitopes detection from amino-acid sequences. I graduated in 2019 with a double degree in Applied Mathematics and Machine Learning from Ecole des Ponts ParisTech and Ecole Normale Supérieure de Paris-Saclay.