Yuyin Zhou

This paper reviews the evolution of deep learning in biomedical image segmentation, highlighting the shift to foundation models and introducing a taxonomy for promptable segmentation methods.

MethylProphet predicts whole-genome DNA methylation without methylation input using a gene-guided Transformer model trained on large-scale bisulfite sequencing data.

EpiFoundation is a foundation model for scATAC-seq that learns cell representations through peak-to-gene alignment, achieving state-of-the-art performance in downstream tasks.

This paper improves gene expression recovery in scRNA-seq by using self-supervised deep learning to model gene-gene interactions.