Xinhao Liu

MOSAICField aligns spatial slices across modalities to reconstruct 3D tissue models and map morphology, outperforming existing methods on simulated and prostate cancer data.

DeST-OT aligns spatiotemporal transcriptomics data from developmental timepoints using optimal transport to model cell growth, death, and differentiation.

This paper introduces HM-OT to infer cell type differentiation maps from developmental transcriptomics time series by simultaneously clustering cells and learning transitions between cell types.

PathOmCLIP infers spatial gene expression from histology images by combining pathology and single-cell RNA-seq foundation models, achieving better gene expression prediction than existing methods.

PASTE2 aligns and integrates multi-slice spatial transcriptomics data with partial overlap, improving alignment accuracy and enabling 3D gene expression reconstruction.