Unveiling Fine-scale Spatial Structures and Amplifying Gene Expression Signals in Ultra-Large ST slices with HERGAST

Abstract

We propose HERGAST, a system for spatial structure identification and signal amplification in ultra-large-scale and ultra-high-resolution spatial transcriptomics data. To handle ultra-large ST data, we consider the divide and conquer strategy and devise a Divide-Iterate-Conque framework specially for spatial transcriptomics data analysis, which can also be adopted by other computational methods for extending to ultra-large-scale ST data analysis. To tackle the potential oversmoothing problem arising from data splitting, we construct a heterogeneous graph network to incorporate both local and global spatial relationships. In simulation, HERGAST consistently outperformed other methods across all settings with more than 10% average gaining. In real-world data, HERGAST's high-precision spatial clustering enabled finding SPP1+ macrophages intermingled in tumors in colorectal cancer, while the enhanced gene expression signal enabled discovering unique spatial expression pattern of key genes in breast cancer.