跟着Nature Biotechnology学作图:R语言ComplexHeatmap包做热图展示TCGA-BRCA数据集概况

论文

Removing unwanted variation from large-scale RNA sequencing data with PRPS

https://www.nature.com/articles/s41587-022-01440-w#data-availability

数据链接

https://zenodo.org/record/6459560#.Y2D2NHZBzid

https://zenodo.org/record/6392171#.Y2D2SXZBzid

代码链接

https://github.com/RMolania/TCGA_PanCancer_UnwantedVariation

今天推文重复的图没有出现在论文中，是论文中提供的代码里的一个图

首先是示例数据集

这个数据集论文中并没有提供，需要运行一系列代码获取，获取这个数据集的代码就不介绍了，推文内容主要介绍是如何绘图

读取数据集

sample.info.01<-read.csv("example_data/sample_info_brca_tcga_rnaseq.csv",
                         row.names = 1)
dim(sample.info.01)
head(sample.info.01)

ComplexHeatmap的帮助文档

https://jokergoo.github.io/ComplexHeatmap-reference/book/

ComplexHeatmap安装

BiocManager::install("ComplexHeatmap")

组图代码

library(ComplexHeatmap)
years.colors <- c(
  'green',
  'purple4',
  'blue',
  'brown',
  'tan1',
  'darkgreen',
  'black'
)
names(years.colors) <- c(
  '2009',
  '2010',
  '2011',
  '2012',
  '2013',
  '2014',
  '2015'
)
H.year <- ComplexHeatmap::Heatmap(
  rev(sample.info.01$year_mda),
  cluster_columns  = FALSE,
  column_names_gp = grid::gpar(fontsize = 12),
  col =  years.colors,
  name = 'Time (years)',
  heatmap_legend_param = list(
    color_bar = "discrete" ,
    ncol = 2,
    title_gp = grid::gpar(fontsize = 12)))
H.year

还可以把好多个热图拼到一起

完整代码

sample.info.01<-read.csv("example_data/sample_info_brca_tcga_rnaseq.csv",
                         row.names = 1)
dim(sample.info.01)
head(sample.info.01)

library(ComplexHeatmap)
years.colors <- c(
  'green',
  'purple4',
  'blue',
  'brown',
  'tan1',
  'darkgreen',
  'black'
)
names(years.colors) <- c(
  '2009',
  '2010',
  '2011',
  '2012',
  '2013',
  '2014',
  '2015'
)
H.year <- ComplexHeatmap::Heatmap(
  rev(sample.info.01$year_mda),
  cluster_columns  = FALSE,
  column_names_gp = grid::gpar(fontsize = 12),
  col =  years.colors,
  name = 'Time (years)',
  heatmap_legend_param = list(
    color_bar = "discrete" ,
    ncol = 2,
    title_gp = grid::gpar(fontsize = 12)))
H.year

n.plate <- length(unique(sample.info$plate_RNAseq)) # 38
colfunc <- grDevices::colorRampPalette(
  RColorBrewer::brewer.pal(11, 'PRGn')[-6])
color.plates <- colfunc(n.plate)
H.plate <- ComplexHeatmap::Heatmap(
  rev(sample.info$plate_RNAseq),
  cluster_rows = FALSE,
  cluster_columns = FALSE,
  column_names_gp = grid::gpar(fontsize = 12),
  col = color.plates,
  name = 'Plates',
  heatmap_legend_param = list(
    color_bar = "discrete" ,
    ncol = 4,
    title_gp = grid::gpar(fontsize = 12)))
### TSS
n.tss <- length(unique(sample.info$tss_RNAseq)) # 40
colfunc <- grDevices::colorRampPalette(
  RColorBrewer::brewer.pal(11, 'BrBG')[-6]
)
color.tss <- colfunc(n.tss)
H.tss <- ComplexHeatmap::Heatmap(
  rev(sample.info$tss_RNAseq),
  cluster_rows = FALSE,
  cluster_columns = FALSE,
  column_names_gp = grid::gpar(fontsize = 12),
  col = color.tss,
  name = 'Tissue source sites',
  heatmap_legend_param = list(
    color_bar = "discrete" ,
    ncol = 4,
    title_gp = grid::gpar(fontsize = 12)))
### Tissue
H.tissue <- ComplexHeatmap::Heatmap(
  rev(sample.info$Tissue.Type),
  cluster_rows = FALSE,
  column_names_gp = grid::gpar(fontsize = 12),
  col = c("#252525", 'blue', "#D9D9D9"),
  name = 'Tissues',
  heatmap_legend_param = list(
    color_bar = "discrete" ,
    direction = "vertical",
    ncol = 1,
    title_gp = grid::gpar(fontsize = 12),
    labels = c(
      'Primary tumor',
      'Metastatic tumor',
      'Adjacent normal')))
### Purity
H.purity <- ComplexHeatmap::Heatmap(
  rev(sample.info$purity_HTseq_FPKM),
  column_names_gp = grid::gpar(fontsize = 12),
  cluster_rows = FALSE,
  name = 'Tumor purity score',
  col = viridis::plasma(n = 10),
  heatmap_legend_param = list(
    title_gp = grid::gpar(fontsize = 12)))
### library size
H.ls <- ComplexHeatmap::Heatmap(
  rev(sample.info$libSize),
  cluster_rows = FALSE,
  name = 'Library size',
  column_names_gp = grid::gpar(fontsize = 12),
  col = viridis::viridis(n = 10),
  heatmap_legend_param = list(
    title_gp = grid::gpar(fontsize = 12)))
### PAM50
H.pam50.tcga <- ComplexHeatmap::Heatmap(
  rev(sample.info$Call),
  cluster_rows = FALSE,
  name = 'PAM50 (TCGA calls)',
  column_names_gp = grid::gpar(fontsize = 12),
  col = pam50.colors,
  heatmap_legend_param = list(
    title_gp = grid::gpar(fontsize = 12)))
### PAM50 genefu
H.pam50.genefu <- ComplexHeatmap::Heatmap(
  rev(sample.info$pam50.geneFu.fpkm),
  cluster_rows = FALSE,
  name = 'PAM50 (Genefu calls)',
  column_names_gp = grid::gpar(fontsize = 12),
  col = pam50.colors,
  heatmap_legend_param = list(
    title_gp = grid::gpar(fontsize = 12)))
### Flow cell chemistry
H.fcch <- ComplexHeatmap::Heatmap(
  rev(sample.info$FcCh),
  cluster_rows = FALSE,
  name = 'Flow cell chemistry',
  column_names_gp = grid::gpar(fontsize = 12),
  col = FcCh.colors,
  heatmap_legend_param = list(
    title_gp = grid::gpar(fontsize = 12),
    direction = "horizontal"))
ComplexHeatmap::draw(
  H.year +
    H.fcch +
    H.plate +
    H.tss +
    H.tissue +
    H.pam50.tcga +
    H.pam50.genefu +
    H.ls +
    H.purity,
  merge_legends = FALSE,
  heatmap_legend_side = 'right')

今天的推文没有对代码的细节进行研究，主要就是能够运行出结果，如果后续需要用到这个R包来作图，可以仔细研究这个R包的函数

示例数据和代码可以给推文点赞 点击在看 最后留言获取

欢迎大家关注我的公众号

小明的数据分析笔记本

小明的数据分析笔记本 公众号 主要分享：1、R语言和python做数据分析和数据可视化的简单小例子；2、园艺植物相关转录组学、基因组学、群体遗传学文献阅读笔记；3、生物信息学入门学习资料及自己的学习笔记！