hello，大家好，今天我們來(lái)分享一下有關(guān)空間轉(zhuǎn)錄組研究腫瘤樣本切片的一些思路糯而，主要的參考文獻(xiàn)是Comprehensive Analysis of Spatial Architecture in Primary Liver Cancer蛛砰，里面的方法都很經(jīng)典芍殖，值得我們多多關(guān)注，關(guān)于文獻(xiàn)的分析內(nèi)容呢外厂，大家感興趣可以看一看冕象，我們這里呢就總結(jié)方法思路，希望大家能用的到汁蝶，我們逐步來(lái)分析
第一部分渐扮、取樣论悴，腫瘤樣本的切片取樣這個(gè)也很有講究，如下圖墓律。（客戶(hù)的切片是不可以展示的膀估，我這里采用了文獻(xiàn)的切片）。
采用切片的思路就是三部分都要取到耻讽，1察纯、正常區(qū)域、2针肥、腫瘤區(qū)域捐寥、3、邊界區(qū)域祖驱，三者必不可少握恳。但是更加建議的是一個(gè)冷凍塊至少切三個(gè)片，分別是捺僻，只有正常區(qū)域的切片乡洼、只有腫瘤區(qū)域的切片和包含正常腫瘤區(qū)域的切片，這樣包含的信息最為全面匕坯。如下圖：
第二部分束昵，空間轉(zhuǎn)錄組的基本分析，這個(gè)地方也需要各位注意葛峻，重點(diǎn)的地方我加粗
For the gene-spot matrixes generated by Space Ranger, some routine statistical analyses were performed firstly, including calculating the number of the detected UMIs (nUMI), and genes (nGene) in each spot. Based on them, the basic quality controls (QC) were applied on the data. In detail, the spots with extremely low nUMI or nGene (outliers), and the spots isolated from the main tissue sections were removed. The genes expressed in less than 3 spots, and mitochondrial, ribosomal genes were filtered. 其實(shí)個(gè)人建議不要去除锹雏，不過(guò)具體情況具體分析，不能干擾下游分析的真實(shí)性术奖。
第三部分礁遵，空間轉(zhuǎn)錄組數(shù)據(jù)的整合分析
After QC, we used the R package harmony (v1.0) (30) to integrate the expression data from different sections of each patient, and used the Seurat package (v3.1.5) to perform the basic downstream analysis and visualization. In detail, we firstly combined the expression matrixes of each patient’s all sections, and performed normalization, log-transformation, centering and scaling on them. Next, we identified 2,000 highly variable genes according to their expression means and variances. Based on them, principal components analysis (PCA) was performed to project the spots into a low-dimensional space, which was defined by the first 20 principal components (PCs). Then, by setting the section source as the batch factor and using the “RunHarmony” function, we iteratively corrected the spots’ low-dimensional PC representation to reduce the of impact of batch effect. After this step, the corrected PC matrixes were used to perform unsupervised shared-nearest-neighbor-based clustering and UMAP (uniform manifold approximation and projection) visualization analysis further. And to compare the clusters at gene level, we identified differentially expressed genes of the all or selected clusters by using fold-change analysis and Wilcoxon Rank Sum test with Bonferroni correction.
第三部分這個(gè)地方大家應(yīng)該都很熟悉了吧，就是單細(xì)胞做harmony矯正的做法采记，這個(gè)地方?jīng)]做過(guò)的面壁反思一下佣耐。
第四部分，Cluster similarity analysis唧龄，這個(gè)也是一個(gè)比較常規(guī)的點(diǎn)兼砖，不過(guò)一般都是10X單細(xì)胞數(shù)據(jù)在用，整合分析之后每個(gè)cluster會(huì)包含不同的樣本既棺，在每個(gè)切片上的空間位置也千差萬(wàn)別讽挟，不過(guò)能聚類(lèi)到一起，說(shuō)明表達(dá)相似丸冕，這里的工作就是比較這些cluster的相關(guān)性耽梅。
For the clusters from different patients, we represented them by their spots’ average expression profiles (the log-transformed normalization values). To reduce the impact of extreme values, we excluded some outlier spots in advance, whose first three PC values beyond the range of the mean±3*standard deviation of the cluster they belonged to. Moreover, only the genes with the mean above 0.1 and the variance above 0.05 across all the cluster expression vectors were retained for the downstream comparison analyses 這個(gè)地方還是很值得注意的，剔除異常值采用的是whose first three PC values beyond the range of the mean±3*standard deviation of the cluster they belonged to晨仑，表示很贊用褐墅。
To measure the clusters’ similarities across patients, we preformed two types of analyses, hierarchical clustering and low-dimensional projection. In detail, we firstly applied PCA on the centered and scaled clusters’ average expression profiles, and used the first five PCs to perform hierarchical clustering拆檬，這里的層次聚類(lèi)采用了前五個(gè)PC，這樣的層次聚類(lèi)大家可以學(xué)一學(xué).
層次聚類(lèi)圖上的信息也很豐富妥凳，colorbar采用的是平均值竟贯。
Besides, the diffusion map was used to project clusters of different patients into a two-dimension space (the first two diffusion components) based on the package destiny (34) with default parameter setting
For convenience of comparison, we annotated each cluster with a region label (normal, stromal, or tumor), which was decided by integrating the information of the cluster’s marker genes and H&E staining images.也是非常好的一個(gè)點(diǎn)，明顯diffusion map 的結(jié)果具有區(qū)域性逝钥，相同的區(qū)域一般聚集在一起屑那。
第五部分，Cell type scoring by a signature-based strategy
At the current Visium ST resolution, each spot may contain approximately 8-20 cells, so that we couldn’t assign a certain cell type for each spot.（這也是限制10X空間轉(zhuǎn)錄組發(fā)展的最大原因）艘款。Considering this, to compare the distribution of cell types across the tissue sections, we proposed a signature-based strategy to score the cell type enrichments in each spot.（marker gene的富集持际，這個(gè)方式我在我的公開(kāi)課上提到過(guò)，marker gene富集的方式看看各個(gè)地方的細(xì)胞類(lèi)型的富集程度）.
做marker gene富集的步驟哗咆，我們來(lái)看一下文章是怎么做的
第一步蜘欲，we curated a set of gene signatures of common cell types in liver cancer based on the Xcell signatures and biology prior knowledge（找marker gene）
第二步，很關(guān)鍵晌柬，Then, we defined the average log-transformed normalization expression values of the genes in the signature as the corresponding cell type scores.（這個(gè)富集分?jǐn)?shù)的計(jì)算方式姥份，讓我猝不及防~~~~~??）。
第三步年碘，Taking advantage of these scores, the cell type relative enrichment degree across different tissue regions can be compared.（嗯澈歉，梯度比較，這個(gè)就比較正常了）屿衅。By testing on some single cell RNA-seq datasets of liver cancer, we proved that our curated gene signatures had high sensitivity and specificity.(marker gene的驗(yàn)證確實(shí)很重要)埃难。
后面作者還進(jìn)行了MIA的分析模式，關(guān)于MIA這里就不展開(kāi)講了涤久，大家可以參考我的文章MIA用于單細(xì)胞和空間的聯(lián)合分析涡尘。which determined the cell type enrichment degrees by performing hypergeometric test on the overlap between the tissue region-specific genes of ST data and the cell type-specific genes of single cell data.
Here, we took advantage of cell type annotation and differential expression gene results of a liver cancer single cell dataset and performed MIA on the clusters of our ST data, so that we can use the p-values of hypergeometric test to measure the enrichment of different cell types in each cluster（下圖C）
By comparing these enrichment degrees and the mean values of our signature-based cell type scores of the all ST clusters, we observed generally high correlation（上圖D）。which proved the reliability of our signature-based cell type scoring method. At the same time, it had the advantage of not requiring single cell data, which was more flexible.
第六部分拴竹，Intratumor spatial heterogeneity measurement 悟衩，衡量空間異質(zhì)性剧罩。兩個(gè)思路transcriptome diversity degree and spatial continuity degree栓拜，我們?cè)敿?xì)看一看。
transcriptome diversity degree,這個(gè)地方有點(diǎn)東西
For the transcriptome diversity degree, we firstly calculated the Pearson correlation coefficients between each pair of tumor region spots based on the highly variable genes.（首先計(jì)算每個(gè)spot的Pearson的相關(guān)性）.然后我們將樣本的轉(zhuǎn)錄組多樣性程度定義為這些相關(guān)性的中值絕對(duì)偏差（MAD）的 1.4826 倍惠昔，這是標(biāo)準(zhǔn)偏差的近似值幕与，但可以避免異常值的影響 。該度量越大意味著樣本腫瘤點(diǎn)之間的相似性具有更大的方差,使樣本具有更高的瘤內(nèi)異質(zhì)性镇防。 公式化地啦鸣，它可以計(jì)算為
where ei indicated the expression vector of the tumor region spot i, and the MAD was defined as
spatial continuity degree
first compared the cluster identities of each tumor region spot with its six neighbor spots
Then the total fraction of the neighbor spots with the same cluster identity was defined as the spatial continuity degree.（然后將具有相同簇標(biāo)識(shí)的相鄰點(diǎn)的總分?jǐn)?shù)定義為空間連續(xù)度。）来氧。該指標(biāo)測(cè)量了腫瘤區(qū)域的空間異質(zhì)性诫给。
The larger this metric meant the sample’s tumor region more tended to be block-like (higher spatial continuity degree and lower spatial mixed degree). Formulaically, it can be calculated as
where i indicated a tumor region spot, and I() was the indicative function.
第七部分香拉，GSVA分析，這部分大家應(yīng)該都知道才對(duì)
In detail, the log-transformed normalization expression matrix of tumor spots was inputted into the “gsva” function with the default parameters setting.
to compare the tumor clusters across patients at pathway level, we averaged the resulting GSVA score matrixes over each cluster and performed hierarchical clustering on them with Ward's minimum variance method （這部分相對(duì)簡(jiǎn)單）中狂。
第八部分凫碌，Spatial gradient change analysis，這個(gè)很重要
The spatial gradient distributions of hallmark pathway activities were analyzed on our leading-edge samples (L-sections) and the intact HCC nodule (HCC-5).
For the leading-edge samples, we focused on analyzing the gradient changes from capsules or tumor-normal boundary lines to the both tumor and normal sides.（正常區(qū)域向腫瘤區(qū)域過(guò)度的地方）胃榕。
we divided the normal and tumor regions into continuous zones parallel to the shape of the boundary lines at intervals of 5 spots（有點(diǎn)意思）盛险。And the gradient changes along these zones were analyzed。
第九部分勋又，空間通訊分析Cluster interaction analysis
這里作者做通訊分析只做臨近c(diǎn)luster的通訊分析苦掘，F(xiàn)or each pair of neighbor tumor clusters, we selected their interface regions with 4 spots wide (2 spots wide for each cluster) and excluded the spots identified as stromal clusters（看來(lái)也不是盲目的全部選擇,體現(xiàn)了空間做通訊位置的重要性）。
方法就是cellphoneDB
第10部分楔壤，Copy number variation (CNV) comparison analysis
作者直接用空間數(shù)據(jù)做inferCNV鹤啡，結(jié)果么，文獻(xiàn)的結(jié)果很符合實(shí)際蹲嚣。
生活很好揉忘，等你超越
