In order to quantify the spatial heterogeneity of tumors at the single cell level, the spatial transcriptomic dataset of breast cancer on the 10×Genomics platform was selected as the study object, and deep graph infomax (DGI) model was used to cluster breast cancer cells. The results showed that the DGI algorithm showed good clustering performance, and the adjusted Rand index reached 0.55. The clustering results were close to manual annotation stratification, the boundary was smooth, and the breast cancer marker genes and differentially expressed genes between cluster 4 and cluster 8 were well identified. The enrichment results showed that these genes were closely related to the occurrence and development of breast cancer. The results of this analysis may provide new insights into the diagnosis and prognosis of breast cancer by identifying markers for clinical diagnosis and treatment of breast cancer patients.
本研究采用一种基于图卷积网络的细胞聚类方法,可以从空间基因表达数据中结合单个细胞的基因表达和复杂的全局空间信息。在DGI聚类之后,进行细胞簇之间基因表达的差异分析、基因本体(Gene ontology,GO)功能富集分析和京都基因与基因组百科全书(Kyoto encyclopedia of genes and genomes,KEGG)通路富集分析,进一步揭示癌症发生、发展的分子机制[9]。本文聚类方法在基于10×Genomics的乳腺癌数据集上进行了测试,实验结果表明,采用DGI模型对空间转录组数据进行聚类,与其他算法相比性能较好,能够发现具有更多区域连续性和更少噪声的区域。通过对乳腺癌数据集进行详细分析,发现肿瘤内部存在更多的异质性,从而为制定靶向治疗策略提供了理论基础[10]。
WUR, GUOW B, QIUX Y, et al. Comprehensive analysis of spatial architecture in primary liver cancer[J]. Science Advances, 2021, 7(51): eabg3750.
[3]
GOHILS H, IORGULESCUJ B, BRAUND A, et al. Applying high-dimensional single-cell technologies to the analysis of cancer immunotherapy[J]. Nature Reviews Clinical Oncology, 2021, 18(4): 244-256.
[4]
HEYDARIA A, SINDIS S. Deep learning in spatial transcriptomics: learning from the next next-generation sequencing[J]. Biophysics Reviews, 2023, 4(1): 011306.
[5]
ZHAOE, STONEM R, RENX, et al. Spatial transcriptomics at subspot resolution with BayesSpace[J]. Nature Biotechnology, 2021, 39(11): 1375-1384.
[6]
PHAMD, TANX, XUJ, et al. stLearn:integrating spatial location,tissue morphology and gene expression to find cell types,cell-cell interacti ons and spatial trajectories within undissociated tissues[DB/OL]. [2024-03-20].
[7]
ZEHRAT, SHAMSM, AHMADZ, et al. Ki-67 quantification in breast cancer by digital imaging AI software and its concordance with manual method[J]. Journal of the College of Physicians and Surgeons-Pakistan: JCPSP, 2023, 33(5): 544-547.
[8]
SUZ Y, NIAZIM K K, TAVOLARAT E, et al. BCR-Net: a deep learning framework to predict breast cancer recurrence from histopathology images[J]. PloS One, 2023, 18(4): e0283562.
[9]
VELIČKOVIĆP, FEDUSW, HAMILTONW L, et al. Deep graph infomax[DB/OL]. [2024-03-20].
[10]
ZHENGB W, FANGL. Spatially resolved transcriptomics provide a new method for cancer research[J]. Journal of Experimental & Clinical Cancer Research, 2022, 41(1): 179.
[11]
WOLFF A, ANGERERP, THEISF J. SCANPY: large-scale single-cell gene expression data analysis[J]. Genome Biology, 2018, 19(1): 15.
WUT Z, HUE Q, XUS B, et al. clusterProfiler 4.0: a universal enrichment tool for interpreting omics data[J]. Innovation, 2021, 2(3): 100141.
[14]
CHANGH J, YANGU C, LAIM Y, et al. High BRCA1 gene expression increases the risk of early distant metastasis in ER+ breast cancers[J]. Scientific Reports, 2022, 12(1): 77.
[15]
HANT, JIANGS L, ZHENGH, et al. Interplay between c-Src and the APC/C co-activator Cdh1 regulates mammary tumorigenesis[J]. Nature Communications, 2019, 10(1): 3716.
XUC, JINX Y, WEIS R, et al. DeepST: identifying spatial domains in spatial transcriptomics by deep learning[J]. Nucleic Acids Research, 2022, 50(22): e131.
[19]
ZHANGZ, ZHENGY L, BIANX W, et al. Identification of key genes and pathways associated with oxidative stress in periodontitis[J]. Oxidative Medicine and Cellular Longevity, 2022, 2022: 9728172.
[20]
DITTMERJ. Biological effects and regulation of IGFBP5 in breast cancer[J]. Frontiers in Endocrinology, 2022, 13: 983793.