Causal Inference
Causal inference is a cornerstone of scientific discovery. We develop theory and methods for identifying optimal interventions and learning causal relationships from data.
Multi-modal Learning
We develop theory and methods for integrating/translating different data modalities and obtaining disentangled representations that elucidate causal relations.
Active Learning & Experimental Design
Given the many possible perturbations in biology (e.g. CRISPR/drugs), we develop robust strategies for iteratively planning, performing, and learning from interventions.
3D Genome Organization and Regulation
The 3D genome organization is intricately linked to gene regulation. We combine sequencing and imaging data with representation learning and causality to elucidate this link.
Cellular Reprogramming
By combining optimal perturbation design (CRISPR/drugs) and causal inference, we seek to develop methods to control cell state transitions in health and disease.
Cells in the Tissue Microenvironment
We integrate rich spatial transcriptomics data to advance our understanding of tissue architecture, cellular dynamics, and disease pathology.
