但是回顾到一篇于2019年5月发表在cell reports杂志的文章:《Dissecting the Single-Cell Transcriptome Network Underlying Gastric Premalignant Lesions and Early Gastric Cancer》,链接是:https://www.sciencedirect.com/science/article/pii/S221112471930525X 它里面的细胞亚群的网络图其实并没有涉及到细胞通讯分析,我看完了全文都没有找到软件工具。
如下所示:
不同细胞亚群的网络图
图例是:(D) The similarity network among diverse epithelial cell types in our dataset.
The thickness of edges in the network was denoted as the Pearson correlation coefficient between the centroids of any pair of cell type
s. See also Figures S15C and S15D.
然后我就看了看 Figures S15C and S15D.如下所示:
网络图的来源
图例是:Related to Figure 6. The similarities between these cell populations were showed in the form of network (c) and correlation matrix (d). The thickness of the edges in the network was denoted as the Pearson correlation coefficient between the centroids of any pair of cell types.
(E) The correlations between the centroids of cancer cells and metaplastic stem-like cells (left) and enterocytes (right).
两个细胞亚群的相关性计算
细胞亚群的相关性计算3个步骤:
First, we assessed the association of each cell type pair by Pearson correlation coefficient between the centroids of them as the thickness of cell type-cell type network.
Then, we derived the marker genes for representative cell types in each lesion by performing differential expression analysis with the threshold as fold change > 1.5 and FDR < 0.01.
Finally, we connected marker genes for with known protein-protein interactions (PPIs) documented in STRING database for each lesion.
一般来说,建议大家直接读原汁原味的综述,比如2020年11月9日,Erick Armingol等在 Nature Reviews Genetics上发表了一篇综述《Deciphering cell–cell interactions and communication from gene expression》,我看到了解读:
从基因表达解读细胞之间的相互作用和交流
,读了一下发现确实超级适合作为细胞通讯分析的背景知识学习材料。