**MarsGT: Multi-omics Analysis for Rare Population Inference using Single-cell Graph Transformer** **Authors:** Xiaoying Wang, Maoteng Duan, Jingxian Li, Anjun Ma, Gang Xin, Dong Xu, Zihai Li, Bingqiang Liu, Qin Ma **Received:** 7 August 2023 **Accepted:** 14 December 2023 **Published online:** 06 January 2024 **Abstract:** Rare cell populations are crucial in neoplastic progression and therapeutic response, offering potential intervention targets. However, their computational identification and analysis often lag behind major cell types. To address this gap, the authors introduce MarsGT, a multi-omics analysis tool for rare population inference using a single-cell graph transformer. MarsGT identifies rare cell populations using a probability-based heterogeneous graph transformer on single-cell multi-omics data. It outperforms existing tools in identifying rare cells across 550 simulated and four real human datasets. In mouse retina data, MarsGT reveals unique subpopulations of rare bipolar cells and a Müller glia cell subpopulation. In human lymph node data, it detects an intermediate B cell population potentially acting as lymphoma precursors. In human melanoma data, it identifies a rare MAIT-like population impacted by a high IFN-I response and reveals the mechanism of immunotherapy. MarsGT offers biological insights and suggests potential strategies for early detection and therapeutic intervention of diseases. **Key Points:** - **MarsGT Framework:** MarsGT is an end-to-end deep learning model that integrates scRNA-seq and scATAC-seq data to identify rare cell populations. - **Heterogeneous Graph Transformer:** It constructs a heterogeneous graph comprising cells, genes, and peaks to facilitate the analysis of multi-omics data. - **Probability-Based Subgraph-Sampling:** MarsGT uses a probability-based subgraph-sampling technique to highlight rare cell-related genes and peaks. - **Performance Evaluation:** MarsGT outperforms existing tools in identifying rare cells across simulated and real datasets, including mouse retina, human lymph node, and melanoma data. - **Biological Insights:** MarsGT uncovers unique subpopulations of rare cells, such as rare bipolar cells in mouse retina and a rare B cell lymphoma population in human lymph node. - **Immunotherapy Strategies:** MarsGT identifies a rare MAIT-like population in melanoma and reveals the mechanism of immunotherapy, suggesting potential strategies for early detection and treatment. **Conclusion:** MarsGT is a powerful tool for identifying rare cell populations and elucidating microenvironmental and immunotherapeutic mechanisms. It sets a promising trajectory for precision medicine by enabling the discovery of disease-associated rare cell populations and uncovering intrinsic regulatory mechanisms that could inform immunotherapy strategies.**MarsGT: Multi-omics Analysis for Rare Population Inference using Single-cell Graph Transformer** **Authors:** Xiaoying Wang, Maoteng Duan, Jingxian Li, Anjun Ma, Gang Xin, Dong Xu, Zihai Li, Bingqiang Liu, Qin Ma **Received:** 7 August 2023 **Accepted:** 14 December 2023 **Published online:** 06 January 2024 **Abstract:** Rare cell populations are crucial in neoplastic progression and therapeutic response, offering potential intervention targets. However, their computational identification and analysis often lag behind major cell types. To address this gap, the authors introduce MarsGT, a multi-omics analysis tool for rare population inference using a single-cell graph transformer. MarsGT identifies rare cell populations using a probability-based heterogeneous graph transformer on single-cell multi-omics data. It outperforms existing tools in identifying rare cells across 550 simulated and four real human datasets. In mouse retina data, MarsGT reveals unique subpopulations of rare bipolar cells and a Müller glia cell subpopulation. In human lymph node data, it detects an intermediate B cell population potentially acting as lymphoma precursors. In human melanoma data, it identifies a rare MAIT-like population impacted by a high IFN-I response and reveals the mechanism of immunotherapy. MarsGT offers biological insights and suggests potential strategies for early detection and therapeutic intervention of diseases. **Key Points:** - **MarsGT Framework:** MarsGT is an end-to-end deep learning model that integrates scRNA-seq and scATAC-seq data to identify rare cell populations. - **Heterogeneous Graph Transformer:** It constructs a heterogeneous graph comprising cells, genes, and peaks to facilitate the analysis of multi-omics data. - **Probability-Based Subgraph-Sampling:** MarsGT uses a probability-based subgraph-sampling technique to highlight rare cell-related genes and peaks. - **Performance Evaluation:** MarsGT outperforms existing tools in identifying rare cells across simulated and real datasets, including mouse retina, human lymph node, and melanoma data. - **Biological Insights:** MarsGT uncovers unique subpopulations of rare cells, such as rare bipolar cells in mouse retina and a rare B cell lymphoma population in human lymph node. - **Immunotherapy Strategies:** MarsGT identifies a rare MAIT-like population in melanoma and reveals the mechanism of immunotherapy, suggesting potential strategies for early detection and treatment. **Conclusion:** MarsGT is a powerful tool for identifying rare cell populations and elucidating microenvironmental and immunotherapeutic mechanisms. It sets a promising trajectory for precision medicine by enabling the discovery of disease-associated rare cell populations and uncovering intrinsic regulatory mechanisms that could inform immunotherapy strategies.