About Me
I work at the interface of machine learning and computational biology, with a focus on perturbation modeling, single-cell genomics, and generative models. Most recently, I was an ML research intern in the Machine Learning Group at Arc Institute, supervised by Yusuf Roohani and Hani Goodarzi. My work focused on understanding how current perturbation models, which have largely been developed on comparatively simple cell-line perturbation datasets, perform on highly heterogeneous in vivo data spanning many diverse and continuous cell types and cell states.
Before Arc, I completed my Master's thesis in Fabian Theis's lab at Helmholtz Munich. I developed a conditional flow-matching model to predict cellular responses to transcription-factor perturbations and identified previously unknown TF combinations for reprogramming fibroblasts into dendritic cells.
I am convinced that the strongest ML systems for biology come from understanding the biological context deeply and tailoring the modeling approach to it, using prior knowledge and structure in the biological data rather than treating the problem as a generic ML task.
Research Experience
ML Research Intern
Machine Learning Group, Arc InstituteEvaluated perturbation models on large-scale in vivo single-cell CRISPR screens, focusing on generalization across heterogeneous cell types and cell states.
Master's Thesis
Theis Lab, Helmholtz Munich“Generative Modeling of Cellular Reprogramming with Flow Matching.” Developed a conditional flow-matching model to predict cellular responses to transcription-factor perturbations and identify previously unknown TF combinations for reprogramming fibroblasts into dendritic cells.