Sheng Yuqi, Wu Jiashuo, Li Xiangmei, Qiu Jiayue, Li Ji, Ge Qinyu, Cheng Liang, Han Junwei
College of Bioinformatics Science and Technology, Harbin Medical University, Harbin 150081, China.
College of School of Biological Science & Medical Engineering, Southeast University, Nanjing 210096, China.
Brief Bioinform. 2023 Mar 19;24(2). doi: 10.1093/bib/bbad074.
Interactions between Tumor microenvironment (TME) cells shape the unique growth environment, sustaining tumor growth and causing the immune escape of tumor cells. Nonetheless, no studies have reported a systematic analysis of cellular interactions in the identification of cancer-related TME cells. Here, we proposed a novel network-based computational method, named as iATMEcell, to identify the abnormal TME cells associated with the biological outcome of interest based on a cell-cell crosstalk network. In the method, iATMEcell first manually collected TME cell types from multiple published studies and obtained their corresponding gene signatures. Then, a weighted cell-cell crosstalk network was constructed in the context of a specific cancer bulk tissue transcriptome data, where the weight between cells reflects both their biological function similarity and the transcriptional dysregulated activities of gene signatures shared by them. Finally, it used a network propagation algorithm to identify significantly dysregulated TME cells. Using the cancer genome atlas (TCGA) Bladder Urothelial Carcinoma training set and two independent validation sets, we illustrated that iATMEcell could identify significant abnormal cells associated with patient survival and immunotherapy response. iATMEcell was further applied to a pan-cancer analysis, which revealed that four common abnormal immune cells play important roles in the patient prognosis across multiple cancer types. Collectively, we demonstrated that iATMEcell could identify potentially abnormal TME cells based on a cell-cell crosstalk network, which provided a new insight into understanding the effect of TME cells in cancer. iATMEcell is developed as an R package, which is freely available on GitHub (https://github.com/hanjunwei-lab/iATMEcell).
肿瘤微环境（TME）细胞之间的相互作用塑造了独特的生长环境，维持肿瘤生长并导致肿瘤细胞的免疫逃逸。然而，尚无研究报道对鉴定癌症相关TME细胞中的细胞相互作用进行系统分析。在此，我们提出了一种基于网络的新型计算方法，名为iATMEcell，以基于细胞-细胞串扰网络鉴定与感兴趣的生物学结果相关的异常TME细胞。在该方法中，iATMEcell首先从多项已发表的研究中手动收集TME细胞类型，并获得其相应的基因特征。然后，在特定癌症大块组织转录组数据的背景下构建加权细胞-细胞串扰网络，其中细胞之间的权重既反映了它们的生物学功能相似性，也反映了它们共享的基因特征的转录失调活性。最后，它使用网络传播算法来鉴定显著失调的TME细胞。使用癌症基因组图谱（TCGA）膀胱尿路上皮癌训练集和两个独立验证集，我们证明iATMEcell可以鉴定与患者生存和免疫治疗反应相关的显著异常细胞。iATMEcell进一步应用于泛癌分析，结果显示四种常见的异常免疫细胞在多种癌症类型的患者预后中起重要作用。总体而言，我们证明iATMEcell可以基于细胞-细胞串扰网络鉴定潜在异常的TME细胞，这为理解TME细胞在癌症中的作用提供了新的见解。iATMEcell作为一个R包开发，可在GitHub（https://github.com/hanjunwei-lab/iATMEcell）上免费获取。
