10X空间转录组在乳腺癌研究中的运用

hello，大家好，本来国庆放假应该好好休息的，很想出去玩，奈何天天下雨，躺下家里也不知道干什么，所以就分享一些内容给大家吧，大家如果跟我一样出不去，不如就好好学习一下，涨涨知识。

今天我们来分享一个利用单细胞空间联合分析深入研究乳腺癌的思路，文章在A single-cell and spatially resolved atlas of human breast cancers,当然，我们这里重点关注空间转录组的研究部分。

当然，文献刚开始对多个乳腺癌单细胞样本进行研究，做了很重要的基础工作，包括细胞定义和CNV识别恶性细胞。

当然，这里作者对细胞类型的识别采用了细胞定义软件Garnett，原理就是构建细胞类型的分类器，不过普适应不强。有关于CNV的识别，主要还是运用到inferCNV，大家应该很熟悉了，不过这里的细节部分大家需要注意一下，Genes with a mean count of less than 0.1 across all cells were filtered out before the analysis and the signal was denoised using a dynamic threshold of 1.3 s.d. from the mean，当然这里最让我惊讶的是，作者这里做inferCNV选取ref的时候，居然是用免疫细胞和内皮细胞用于定义参考细胞推断的拷贝数分布，Epithelial cells were used for the observations，这个地方还是本人第一次遇见，之前一直都不太认同。Briefly, inferred changes at each genomic locus were scaled (between −1 and +1) and the mean of the squares of these values was used to define a genomic instability score for each cell，这里有一个不稳定性的评分，看来跟之前分享的信噪比内容相似，In each individual tumor, the top 5% of cells with the highest genomic instability scores were used to create an average CNV profile。Each cell was then correlated to this profile.Cells were plotted with respect to both their genomic instability and correlation scores. Partitioning around medoids clustering was performed using the pamk function in the R package cluster v.2.0.7-1 to choose the optimum value for k (between 2 and 4) using silhouette scores and the pam function to apply the clustering. Thresholds defining normal and neoplastic cells were set at 2 cluster s.d. to the left and 1.5 s.d. below the first cancer cluster means. For tumors where partitioning around medoids could not define more than 1 cluster, the thresholds were set at 1 s.d. to the left and 1.25 s.d. below the cluster means. （对细胞的基因组不稳定性和相关性评分进行绘图。 使用 R 包 cluster v.2.0.7-1 中的 pamk 函数执行围绕 medoids 聚类的分区，以使用轮廓分数和 pam 函数选择 k 的最佳值（2 到 4 之间）以应用聚类。 定义正常和肿瘤细胞的阈值设置为 2 簇 s.d。这个地方还是需要注意，定义的算是很精细了）。看来肿瘤个体间的异质性很大。This revealed patient-unique copy number changes and those commonly seen in breast cancers, such as chr1q gain in luminal cancers and chr5q loss in basal-like breast cancers。

当然，单细胞部分接下来就是对大类进行细分，这个在单细胞分析中也很常见。当然这里涉及到一个共享cluster分析，大家可以参考cola

当然为了更加准确，作者自己采用了一种叫SCSubtype的方法，个性化程度很高，普适性不强。当然，后面对肿瘤内部异质性的研究方法也很常规，再分群 + 功能富集

然后对基因模块进行打分（这里的打分作者使用的是AUcell），分析肿瘤的异质性。

后续对肿瘤的免疫微环境进行研究，

当然，也做了通讯分析（这个图是基本功，我希望看到的分析人员可以自己画出来）

当然，单细胞的轨迹分析揭示了不同的细胞分化状态（对识别肿瘤等级很有帮助，但是轨迹分析远没有看起来的那么简单）。

接下来就是空间转录组了，当然，主要也是单细胞空间进行联合（方法是Stereoscope，文献在Single-cell and spatial transcriptomics enables probabilistic inference of cell type topography，感兴趣大家可以看看）。

当然，首先是对单细胞分析得到的gene module进行富集打分分析（AUcell，分析模块的空间和个体情况）

接下来就是联合了，看看细胞类型的空间分布

其中最为关键是，就是cell-cell的interaction，也就是说一种细胞类型的周围富集了哪些细胞类型，类似于之前分享的细胞单元，大家可以参考文章10X空间转录组数据分析之细胞单元，For cell–cell colocalizations across all tissue domains, we included seven additional HER2+ datasets generated on a platform similar to Visium56. In total, Pearson correlation was computed from the cell abundances across the tissue locations from 13 patients using R (cor.test function; cutoff P = 0.05). For cell signaling predictions between iCAFs and CD4/CD8+ T cells, spots containing the two cell types of interest were first selected using the product of the two respective deconvolution values. Interaction scores were defined as the product of the ligand and receptor log expression levels using two independent cell signaling sets70,71 and only ligands and receptors differentially expressed by iCAFs and CD4/CD8+ T cells in the scRNA-seq data, respectively (MAST; average log fold change threshold = 0.1). All regions annotated as normal ductal by pathology were also excluded from the above analyses.（细胞类型的共定位分析非常重要，希望做相关研究的人员可以很好的重视）。

生活很好，有你更好