将方面合并到一个循环中

Question

我有以下数据集（数据集太大而无法使用 dput 所以现在只使用 head() ）：

head(df3_melt_mean)
  Isodecoder Anticodon HEK293_rep1_smallRNAseq HEK293_rep2_smallRNAseq HEK293_rep3_smallRNAseq HEK293_rep1_hydroseq HEK293_rep2_hydroseq
1        Ala       AGC            24442.262951            22452.241238            22390.900942           9971.50351          11438.44236
2        Ala       CGC              743.402305              785.417884              785.435272           3493.42947           3757.57860
3        Ala       GGC                1.058978                2.380054                1.498922             20.98667             13.54833
4        Ala       TGC             3507.333950             3755.725518             3613.901605           7386.74006           7681.38485
5        Arg       GCG                0.000000                0.000000                0.000000              0.00000              0.00000
6        Arg       ACG             2821.116438             2731.112187             2856.945856          10353.23399           9125.32470
  HEK293_rep3_hydroseq Pan_tRNA1 Pan_tRNA2     HEK1      HEK2     HEK3     HEK4   Label variable        value          Condition
1          11213.43885 35665.006 32012.831 22247.59  29870.42 27672.74 35719.69 Ala-AGC ARM_mean 23095.135043 ARM-seq - Mean CPM
2           3557.33569  9930.173  7260.746 14897.03  14413.76 14970.47 12177.89 Ala-CGC ARM_mean   771.418487 ARM-seq - Mean CPM
3             18.69229     0.000     0.000     0.00      0.00     0.00     0.00 Ala-GGC ARM_mean     1.645985 ARM-seq - Mean CPM
4           7567.79822 10365.656  8080.036 35228.45  42094.44 34821.68 29162.91 Ala-TGC ARM_mean  3625.653691 ARM-seq - Mean CPM
5              0.00000     0.000     0.000     0.00      0.00     0.00     0.00 Arg-GCG ARM_mean     0.000000 ARM-seq - Mean CPM
6          10847.97240 11070.480 12520.278 81727.21 108113.40 62025.37 63240.17 Arg-ACG ARM_mean  2803.058160 ARM-seq - Mean CPM

我写了一个循环，它将 plot 分别归一化每个氨基酸的计数，但我正在努力编写一个循环，将所有 22 个图包装成一个大平面包装。 理想情况下，我想拥有所有单独的 22 个地块，但我也想拥有一个 plot，其中包含所有 22 个地块（用于演示目的）。 我的循环如下：

 for (iso in c("Ala", "Arg", "Asn", "Asp", "Cys", "Gln", "Glu", "Gly", "His", "Ile", "Leu", "Lys", "Met", "Phe", "Pro", "SeC", "Ser", "Thr", "Trp", "Tyr", "Val", "iMet")) {
  
  df4 <- subset(df3_melt_mean, Isodecoder == iso) 
  
  normalized_counts_graphs <- ggplot(data = df4, aes(x = Label, y = value, colour = Condition, group = Condition))+
    geom_point(position = position_dodge(width = 0.2), aes(size = 2))+
    xlab("Isodecoder")+
    ylab("Normalized Counts (CPM)")+
    theme_bw()+
    theme(axis.text.x = element_text(size = 12, angle = 45, hjust = 1), plot.margin = margin(0.5,0.5,0.5,2, "cm"), axis.title = element_text(size = 12))+
    ggtitle(iso, "Comaprison of CPM's for different tRNA-seq methodologies")+
    ggsave(filename = paste(iso, "_methodsComparison_normalizedcounts.png", sep = "_"), bg = "white", width = 15, height = 7, dpi = 700)
}

提前致谢！

Answer 1

我建议使用您提供的数据并在您的代码中添加一些关于facet_wrap()的详细信息，遵循伟大的 @AllanCameron 和 @starja 的评论的下一种方法：

数据：

df <- structure(list(Isodecoder = c("Ala", "Ala", "Ala", "Ala", "Arg", 
"Arg"), Anticodon = c("AGC", "CGC", "GGC", "TGC", "GCG", "ACG"
), HEK293_rep1_smallRNAseq = c(24442.26295, 743.402305, 1.058978, 
3507.33395, 0, 2821.116438), HEK293_rep2_smallRNAseq = c(22452.24124, 
785.417884, 2.380054, 3755.725518, 0, 2731.112187), HEK293_rep3_smallRNAseq = c(22390.90094, 
785.435272, 1.498922, 3613.901605, 0, 2856.945856), HEK293_rep1_hydroseq = c(9971.50351, 
3493.42947, 20.98667, 7386.74006, 0, 10353.23399), HEK293_rep2_hydroseq = c(11438.44236, 
3757.5786, 13.54833, 7681.38485, 0, 9125.3247), HEK293_rep3_hydroseq = c(11213.43885, 
3557.33569, 18.69229, 7567.79822, 0, 10847.9724), Pan_tRNA1 = c(35665.006, 
9930.173, 0, 10365.656, 0, 11070.48), Pan_tRNA2 = c(32012.831, 
7260.746, 0, 8080.036, 0, 12520.278), HEK1 = c(22247.59, 14897.03, 
0, 35228.45, 0, 81727.21), HEK2 = c(29870.42, 14413.76, 0, 42094.44, 
0, 108113.4), HEK3 = c(27672.74, 14970.47, 0, 34821.68, 0, 62025.37
), HEK4 = c(35719.69, 12177.89, 0, 29162.91, 0, 63240.17), Label = c("Ala-AGC", 
"Ala-CGC", "Ala-GGC", "Ala-TGC", "Arg-GCG", "Arg-ACG"), variable = c("ARM_mean", 
"ARM_mean", "ARM_mean", "ARM_mean", "ARM_mean", "ARM_mean"), 
    value = c(23095.13504, 771.418487, 1.645985, 3625.653691, 
    0, 2803.05816), Condition = c("ARM-seq-MeanCPM", "ARM-seq-MeanCPM", 
    "ARM-seq-MeanCPM", "ARM-seq-MeanCPM", "ARM-seq-MeanCPM", 
    "ARM-seq-MeanCPM")), class = "data.frame", row.names = c(NA, 
-6L)) 

在单个图的第一种情况下，这里的代码对换行进行了轻微修改：

library(ggplot2)
#Loop for individuals plots
for (iso in unique(df$Isodecoder)) {

df4 <- subset(df, Isodecoder == iso) 

normalized_counts_graphs <- ggplot(data = df4, aes(x = Label, y = value, colour = Condition, group = Condition))+
  geom_point(position = position_dodge(width = 0.2), aes(size = 2))+
  facet_wrap(.~Isodecoder)+
  xlab("Isodecoder")+
  ylab("Normalized Counts (CPM)")+
  theme_bw()+
  theme(axis.text.x = element_text(size = 12, angle = 45, hjust = 1), plot.margin = margin(0.5,0.5,0.5,2, "cm"), axis.title = element_text(size = 12))+
  ggtitle(iso, "Comaprison of CPM's for different tRNA-seq methodologies")+
  ggsave(filename = paste(iso, "_methodsComparison_normalizedcounts.png", sep = "_"), bg = "white", width = 15, height = 7, dpi = 700)
}

这将为您的每个代码创建一个单独的 plot。 结果如下：

最后，全局 plot 需要这样的代码：

#General plot
ggplot(data = df, aes(x = Label, y = value, colour = Condition, group = Condition))+
  geom_point(position = position_dodge(width = 0.2), aes(size = 2))+
  facet_wrap(.~Isodecoder,ncol = 2)+
  xlab("Isodecoder")+
  ylab("Normalized Counts (CPM)")+
  theme_bw()+
  theme(axis.text.x = element_text(size = 12, angle = 45, hjust = 1),
        plot.margin = margin(0.5,0.5,0.5,2, "cm"),
        axis.title = element_text(size = 12))+
  ggtitle("Comaprison of CPM's for different tRNA-seq methodologies")+
  ggsave(filename ="methodsComparison_normalizedcounts.png",
         bg = "white", width = 15, height = 7, dpi = 700)

Output：

您可以在facet_wrap()中使用ncol调整列数。