Journal Article

SpaSEG: unsupervised deep learning for multi-task analysis of spatially resolved transcriptomics

Chen, Ao, Wei, Yilin, Xu, Xun, Zheng, Bingjie, Zhang, Wenxi, Luo, Qiuhong, Bai, Yong, Jin, Xin, Wang, Yingyue, Wu, Liang, Guo, Xiangyu, Li, Yuxiang, Liu, Chuanyu, Yin, Jianhua, Wang, Xiangdong, Liu, Keyin, Zhang, Yong

2025

Overview

Spatially resolved transcriptomics (SRT) for characterizing spatial cellular heterogeneities in tissue environments requires systematic analytical approaches to elucidate gene expression variations within their physiological context. Here, we introduce SpaSEG, an unsupervised deep learning model utilizing convolutional neural networks for multiple SRT analysis tasks. Extensive evaluations across diverse SRT datasets generated by various platforms demonstrate SpaSEG’s superior robustness and efficiency compared to existing methods. In the application analysis of invasive ductal carcinoma, SpaSEG successfully unravels intratumoral heterogeneity and delivers insights into immunoregulatory mechanisms. These results highlight SpaSEG’s substantial potential for exploring tissue architectures and pathological biology.

• Publisher: BioMed Central,Springer Nature B.V,BMC
• Subject: Animal Genetics and Genomics / Bioinformatics / Biomedical and Life Sciences / Breast Neoplasms - genetics / carcinoma / Carcinoma, Ductal, Breast - genetics / Cell–cell interaction / Clustering / Correspondence / Data analysis / data collection / Datasets / Deep Learning / Evolutionary Biology / Female / Gene expression / Gene Expression Profiling - methods / genome / Human Genetics / Humans / Hypothesis testing / Identification / immunomodulation / Immunoregulation / Life Sciences / Method / Microbial Genetics and Genomics / Multi-section integration / Neural networks / Plant Genetics and Genomics / Spatial domain identification / Spatially resolved transcriptomics / Spatially variable gene / Task analysis / Transcriptome / Transcriptomics / Unsupervised Machine Learning