Xiucai Ye

The paper links dietary patterns with serum Raman spectra and metabolomics in Japanese adults, showing dietary information is distributed across complementary phenotypic layers.

This study evaluates the zero-shot performance of single-cell foundation models across tasks and datasets, showing task-dependent utility and structural robustness vary with dataset characteristics.

DiffusionST improves spatial transcriptomics data quality and identifies spatial domains through a diffusion model and graph convolutional network.

DenoiseST is a deep learning method for spatial transcriptomics that improves data imputation, clustering, and identification of functionally variable genes by reducing dropout noise.

This paper introduces CSDGI, a method to infer cancer subtype-specific driver genes from single-cell transcriptomics data, showing it effectively identifies genes associated with key biological processes and disease pathways.