Xinyi Zhang

This paper introduces a decomposition framework for perturbation response prediction, showing that aligning models with biological priors improves generalization across unseen perturbations and cell lines.

APOLLO learns shared and modality-specific information from multi-modal single-cell data to improve understanding of cell state and regulatory mechanisms.

This paper introduces PUPS, a method that predicts protein subcellular localization in single cells using protein sequences and cellular images, showing generalization to unseen proteins and cell lines.