Minxing Pang

TissueCraftAI generates multi-modal spatial omics data from histology images using natural language prompts, outperforming existing methods in image generation and molecular data prediction.

This study introduces VAE-PRS, a deep learning model that improves polygenic prediction for blood cell traits by capturing interaction effects and outperforming traditional methods in 14 out of 16 traits.

CelloType is a multi-task learning model for cell segmentation and classification in tissue images, improving accuracy over existing methods.

This study maps human bone marrow niches using single-cell transcriptomics and proteomic imaging, identifying distinct cellular environments and their roles in hematopoiesis and AML.

Modelling all spots of a section jointly with a Vision Transformer, HisToGene predicts expression from H&E images better than a re-implemented ST-Net and adds 4x super-resolution prediction.

The paper introduces a multitask learning module for Transformers to analyze large-scale single-cell transcriptomes, achieving better performance than autoencoder and PCA on the Mouse Brain Atlas.