Interactions: We developed a method for quantifying molecular interactions using stochastic modelling and super-resolution microscopy (Bermudez-Hernandez & Keegan et al., Scientific Reports 2017).

Deep learning applied to histopathology: We apply deep learning methods to analyze tissue images and to integrate multi-omic data. We have developed models to predict endometrial cancer subtypes and molecular features from histopathology images using a multi-resolution models (Hong et al., Cell Reports Medicine 2021), and to integrate histopathology and proteogenomics at a pan-cancer level (Wang et al., Cell Reports Medicine 2023)