基于Seurat的空转单样本数据分析流程学习(六)-CNV分析

原创

凑齐六个字吧

发布于 2025-10-12 22:09:20

13600

代码可运行

文章被收录于专栏：单细胞单细胞

运行总次数：0

代码可运行

虽然这里说是基于Seurat，但其实主要还是把Seurat对象转化成了SPATA2这个包进行了分析。SPATA2是一个专门针对空间转录组数据（spatial transcriptomics）分析的R包，主要用于处理10x Genomics Visium或其他空间转录组技术的数据。它整合了单细胞转录组分析方法和空间信息，使用户可以在组织切片上可视化基因表达、细胞类型分布和空间模式。这个R包把inferCNV的功能给封装进来，使用起来更加方便，因此可以用这个包去做空转的CNV分析。

分析流程

1.导入

rm(list = ls())
library(SPATA2)
library(tidyverse)
library(Seurat)
library(future)
library(ggplot2)
library(hdf5r)
library(qs)
library(tidyverse)

sce <- qread("./sce.all.qs")

# check数据
SpatialFeaturePlot(sce, 
                   features = c("T", "fibroblasts"), 
                   pt.size.factor = 3, 
                   ncol = 2, 
                   crop = TRUE)

2.Seurat转SPATA2

# use Spatial assay
sce_SPATA2 <-  asSPATA2(object = sce,
                           sample_name ="GSM8633891",
                           platform ="VisiumSmall",
                           img_scale_fct ="lowres",
                           assay_name ="Spatial",
                           assay_modality ="gene" )

plotSurface(object = sce_SPATA2, color_by = "second_type")
# plotSurface(object = sce_SPATA2, color_by = "seurat_clusters")

3.CNV分析

dir.create("OutPut")
# change input
sce_SPATA2 <- 
  runCNV(
    object = sce_SPATA2,
    directory_cnv_folder = "OutPut", 
    clear_noise_via_ref_mean_sd = list(sd_amplifier = 2)
    )

4.可视化

4.1 基础可视化

cnv_results <- getCnvResults(sce_SPATA2)
names(cnv_results)
plotCnvHeatmap(object = sce_SPATA2, 
               across = "second_type", 
               clrp = "npg")

# a more complex set up
plotCnvHeatmap(
  object = sce_SPATA2, 
  across = "second_type", 
  across_subset = c("malignant", "T"), # dont show the transition part
  meta_vars = "CD4",  # visualize CD4 expression on the left
  meta_vars_clrs = c("CD4" = "inferno"), # with the inferno color spectrum 
  chrom_subset = c("1","2","3","4","5",
                   "6","7","8","9","10", 
                   "11","12","13","14","15",
                   "16","17","18","19","20",
                   "21","22"), # only show these chromosomes
  ggpLayers = list(arm = list(legendNone())) # remove the chrom arm legend
  )

# 基因绘图
plotSurface(sce_SPATA2, 
            color_by = "CD4", 
            pt_clrsp = "inferno")

但结果似乎并不能很好的区分肿瘤细胞和其他细胞，个人认为是跟之前的注释和SPOT的缺陷性有关。

4.2 折线图

plotCnvLineplot(
  object = sce_SPATA2,
  across = "second_type", 
  n_bins_bcsp = 1000,
  n_bins_genes = 1000,
  nrow = 3
)

没啥差别

4.3 Surface展示

每条染色体的拷贝数变异所对应的数值，会被立即写入SPATA2 对象的特征数据中，因此这些数值可以被所有支持数值型变量的函数直接访问。在函数中指定的参数cnv_prefix会与染色体编号组合，形成这些特征变量的名称，用于后续引用。

# cnv feature names
getCnvFeatureNames(object = sce_SPATA2) %>% head()

# are part of all feature names
getFeatureNames(object = sce_SPATA2) %>% head()

plotSurface(
  object = sce_SPATA2, 
  color_by = "Chr7", 
  pt_clrsp = "Reds"
)

plotSurface(
  object = sce_SPATA2, 
  color_by = "Chr10", 
  pt_clrsp = "Oslo" 
)

4.4 Gradients展示

plotSpatialTrajectories(
  object = sce_SPATA2, 
  ids = "horizontal_mid", 
  color_by = "Chr7", 
  pt_clrsp = "Reds 3"
)

plotSpatialTrajectories(
  object = sce_SPATA2, 
  ids = "horizontal_mid", 
  color_by = "Chr10", 
  pt_clrsp = "Oslo"
)

plotStsLineplot(
  object = sce_SPATA2, 
  id = "horizontal_mid", 
  variables = "Chr7" 
)

plotStsLineplot(
  object = sce_SPATA2, 
  id = "horizontal_mid", 
  variables = "Chr10", 
  line_color = "blue" 
)

参考资料：

Inferring histology-associated gene expression gradients in spatial transcriptomic studies. Nat Commun. 2024 Aug 23;15(1):7280.
SPATA2 github:https://github.com/theMILOlab/SPATA2

注：若对内容有疑惑或者有发现明确错误的朋友，请联系后台。更多相关内容可关注公众号：生信方舟 。

原创声明：本文系作者授权腾讯云开发者社区发表，未经许可，不得转载。

如有侵权，请联系 cloudcommunity@tencent.com 删除。

数据挖掘

原创声明：本文系作者授权腾讯云开发者社区发表，未经许可，不得转载。

如有侵权，请联系 cloudcommunity@tencent.com 删除。

数据挖掘

#单细胞

#空间转录组

登录后参与评论

0 条评论

热度