INSTINCT: Multi-sample integration of spatial chromatin accessibility sequencing data via stochastic domain translation

• DOI: 10.1038/s41467-025-56535-0
• Publication Date: 2025-02-01T14:03:05Z
• Authors: Yuyao Liu; Zhen Li; Xiaoyang Chen; Xuejian Cui; Zijing Gao; Rui Jiang
• Keywords: Sample (material); Instinct

Recent advances in spatial epigenomic techniques have given rise to assay for transposase-accessible chromatin using sequencing (spATAC-seq) data, enabling the characterization of heterogeneity and information simultaneously. Integrative analysis multiple spATAC-seq samples, which no method has been developed, allows effective identification elimination unwanted non-biological factors within comprehensive exploration tissue structures providing a holistic landscape, thereby facilitating discovery biological implications study regulatory processes. In this article, we present INSTINCT, multi-sample INtegration Spatial chromaTIN accessibility data via stochastiC domain Translation. INSTINCT can efficiently handle high dimensionality eliminate complex noise batch effects samples through stochastic translation procedure. We demonstrate superiority robustness integrating across simulated scenarios real datasets. Additionally, highlight advantages identification, visualization, spot-type annotation, various downstream analyses, including motif enrichment analysis, expression partitioned heritability analysis. Complex between hinder their joint Here, authors integration. They show that effectively remove while preserving sufficient variations tasks.