如何修复scale_colour_manual（）和geom_point（）的大小固定数目N的范围

Question

我有以下的数据帧：

library(tidyverse) 
dat <- structure(list(motif_and_gene_name = c("FCCT", "XXX4", "XXX2", 
"FCCT", "XXX4", "XXX2", "FCCT", "XXX4", "XXX2", "FCCT", "XXX4", 
"XXX2", "FCCT", "XXX4", "XXX2", "FCCT", "XXX4", "XXX2"), 
    tissue = c("DA", "DA", "DA", "MB", 
    "MB", "MB", "VL", "VL", "VL", "UP", 
    "UP", "UP", "BU", "BU", "BU", 
    "TV", "TV", "TV"), motif_enrichment = c(4740, 
    964.2, 539.2, 6634, 1860, 1150, 6312, 2146, 1432, 5336, 2282, 
    1381, 2796, 1947, 1175, 8190, 1576, 926.8), expression_tpm = c(5.095, 
    15.1825, 1.4225, 7.27, 23.7125, 6.85, 4.8775, 27.17, 3.0025, 
    6.0025, 23.3725, 5.1425, 5.4525, 20.215, 4.695, 6.44, 22.04, 
    4.24)), class = c("tbl_df", "tbl", "data.frame"), row.names = c(NA, 
-18L), .Names = c("motif_and_gene_name", "tissue", "motif_enrichment", 
"expression_tpm")) 

dat 
#> # A tibble: 18 x 4 
#> motif_and_gene_name tissue motif_enrichment expression_tpm 
#>     <chr> <chr>   <dbl>   <dbl> 
#> 1    FCCT  DA   4740.0   5.0950 
#> 2    XXX4  DA   964.2  15.1825 
#> 3    XXX2  DA   539.2   1.4225 
#> 4    FCCT  MB   6634.0   7.2700 
#> 5    XXX4  MB   1860.0  23.7125 
#> 6    XXX2  MB   1150.0   6.8500 
#> 7    FCCT  VL   6312.0   4.8775 
#> 8    XXX4  VL   2146.0  27.1700 
#> 9    XXX2  VL   1432.0   3.0025 
#> 10    FCCT  UP   5336.0   6.0025 
#> 11    XXX4  UP   2282.0  23.3725 
#> 12    XXX2  UP   1381.0   5.1425 
#> 13    FCCT  BU   2796.0   5.4525 
#> 14    XXX4  BU   1947.0  20.2150 
#> 15    XXX2  BU   1175.0   4.6950 
#> 16    FCCT  TV   8190.0   6.4400 
#> 17    XXX4  TV   1576.0  22.0400 
#> 18    XXX2  TV   926.8   4.2400 

用下面的代码

mycol <- c("#3D1E1F","#FE677E","#F19068","#E8A99D","#FCDED4") 
nof_vals <- length(dat$expression_tpm) 

dat %>% 
    ggplot(aes(x=tissue,y=motif_and_gene_name, size=motif_enrichment, colour=factor(expression_tpm))) + 
    geom_point(stat = "identity") + 
    scale_colour_manual(values = colorRampPalette(mycol)(nof_vals)) + 
    theme_minimal() 

我可以创建这个：

圆的大小由motif enrichment决定，颜色范围是expression tpm。注意现在有4种尺寸的圆形和18种颜色。如何将尺寸固定到N = 5的尺寸和颜色，并使用预变形的断点：

size range: motif_enrichment <=100 101-500 501-1000 1001-1500 >1500 
color range: expression_tpm <=1 2-5  6-10  11-15 >15 

Answer 1

您必须创建一个变量来定义断点。使用cut。

dat$expr_brks <- cut(dat$expression_tpm, breaks = c(-Inf, 1, 5, 10, 16, 20, Inf)) 
dat$motif_brks <- cut(dat$motif_enrichment, breaks = c(-Inf, 100, 500, 1000, 1500, 2000, Inf)) 

dat %>% 
    ggplot(aes(x = tissue,y = motif_and_gene_name, size = motif_brks, color = expr_brks)) + 
    geom_point(stat = "identity") + 
    scale_color_manual(values = colorRampPalette(mycol)(length(levels(dat$expr_brks)))) + 
    theme_minimal()