重铸数据框并选择要删除的值的顺序

Question

我有以下数据框。我想为一个变量重铸数据框，并为每个单元格保留一个值。该值不仅是第一个，而且我想使用手动设置放置顺序。

例如： Variant_Classification.recode 我想按 SNV 的降序排列 Indel Splice_Site Translation_Start_Site 和最后一个 Noncoding。

我苦苦挣扎了一会儿，但我没有找到任何解决方案！

谢谢！

mydata.df.test_3_mutations_col  <- structure(list(Individual = structure(c(1L, 1L, 1L, 3L, 3L, 3L, 3L, 3L, 3L, 5L, 5L, 5L, 6L, 6L,6L, 6L, 8L, 8L, 8L, 8L), .Label = c("p1", "p2", "p3", "p4", "p5", "p6", "p7", "p8", "p9"), class = "factor"), Variant_Classification.recode = structure(c(3L, 1L, 2L, 4L, 5L, 3L, 4L, 2L, 1L, 3L, 2L, 2L, 3L, 3L, 5L, 1L, 3L,5L, 1L, 3L), .Label = c("Noncoding", "Indel", "SNVs", "Splice_Site", "Translation_Start_Site"), class = "factor"), Canonical_Hugo_Symbol = structure(c(1L,1L, 1L, 1L, 1L, 1L, 1L, 2L, 2L, 2L, 2L, 2L, 2L, 1L, 1L, 1L, 1L,1L, 1L, 1L), .Label = c("Gene1", "Gene2"), class = "factor")), .Names = c("Individual", "Variant_Classification.recode","Canonical_Hugo_Symbol"), row.names = c(50L, 51L, 116L, 166L, 167L, 168L, 169L, 292L, 293L, 342L, 343L, 344L, 357L, 358L, 359L, 462L, 522L, 523L, 524L, 631L), class = "data.frame")

mydata.reshape <- reshape(mydata.df.test_3_mutations_col, direction = 'wide', idvar = 'Individual', timevar = 'Canonical_Hugo_Symbol')
 colnames(mydata.reshape) <- sub("Variant_Classification.recode.(.*?)","\\1",    colnames(mydata.reshape))
 df <- melt(mydata.reshape,  id = 'Individual', variable_name = 'Canonical_Hugo_Symbol')
 df <- na.omit(df)

q <- ggplot(df,aes(Individual,variable,fill=value))+geom_raster() # Plot mutation group and fill with gene symbol
q + theme(axis.text.x=element_text(angle=-90, hjust = 1, size = 4), text = element_text(size=8), panel.background =  element_blank() , panel.grid.major =  element_line(colour = "black", size = 0.03)) + labs(x = "Patients ID", y = "Gene Symbol") + scale_fill_brewer(palette="Set1", name = "Variant Classification")

现在我找到了解决办法！

首先我必须重新排序因子级别和数据框

mydata.df.test_3_mutations_col$Variant_Classification.recode  <- factor(mydata.df.test_3_mutations_col$Variant_Classification.recode, levels = c("SNVs", "Indel",  "Splice_Site", "Translation_Start_Site","Noncoding"),ordered = TRUE)
mydata.df.test_3_mutations_col  <- mydata.df.test_3_mutations_col[order(mydata.df.test_3_mutations_col$Variant_Classification.recode),]

然后我可以为 ggplot 进行重塑！

mydata.reshape <- reshape(mydata.df.test_3_mutations_col, direction = 'wide', idvar = 'Individual', timevar = 'Canonical_Hugo_Symbol')
   colnames(mydata.reshape) <- sub("Variant_Classification.recode.(.*?)","\\1",       colnames(mydata.reshape))
    df <- melt(mydata.reshape,  id = 'Individual', variable_name = 'Canonical_Hugo_Symbol')
    df <- na.omit(df)

Answer 1

您可以尝试reshape2 包中的dcast，您可以在其中指定自定义聚合函数以根据您的喜好选择值，就像我们在此处所做的那样：

pref.order <- c("SNVs", "Indel", "Splice_Site", "Translation_Start_Site", "Noncoding")
my.fun <- function(x) 
  if(length(x)) 
    head(as.character(x[order(match(x, pref.order))]), 1L) else NA_character_ 

library(reshape2)
dcast(
  mydata.df.test_3_mutations_col, 
  Individual ~ Canonical_Hugo_Symbol, 
  value.var="Variant_Classification.recode",
  fun.aggregate=my.fun
)

生产：

  Individual Gene1 Gene2
1         p1  SNVs  <NA>
2         p3  SNVs Indel
3         p5  <NA>  SNVs
4         p6  SNVs  SNVs
5         p8  SNVs  <NA>