在差异表达值的数据框架中，对每组显著上调和下调的基因进行计数。显著性应通过FDR（假发现率=来自Benjamini的调整p值）和倍数变化来定义。结果应为每组上调和下调的曲线图。（甜蜜的奖励：在图中显示不同的Fc水平（例如：0.5，1，2，4，〉4）.我的解决方案似乎太复杂了，一定有更简单的方法。

示例数据

# create dex df
gene_creator <- paste("gene",1:1000,sep="")
genes = sample(gene_creator,100)

dex_A <- data.frame(
  gene = genes,
  group = "group_A",
  logFC = sample(c(-5:5), replace=T, size=100),
  FDR = sample(c(0.01,1), replace=T, size=100)
)

dex_B <- data.frame(
  gene = genes,
  group = "group_B",
  logFC = sample(c(-5:5), replace=T, size=100),
  FDR = sample(c(0.01,1), replace=T, size=100)
)

dex_C <- data.frame(
  gene = genes,
  group = "group_C",
  logFC = sample(c(-5:5), replace=T, size=100),
  FDR = sample(c(0.01,1), replace=T, size=100)
)

dex_D <- data.frame(
  gene = genes,
  group = "group_D",
  logFC = sample(c(-5:5), replace=T, size=100),
  FDR = sample(c(0.01,1), replace=T, size=100)
)

dex_df <- rbind(dex_A, dex_B, dex_C, dex_D)

溶液

library("tidyverse")

# FC up
dex_up <- dex_df %>% 
  group_by(group) %>%
  filter(FDR <= 0.05) %>% 
  filter(logFC > 0.5 ) %>%
  summarise(n_up = n())

# Fc down
dex_down <- dex_df %>% 
  group_by(group) %>%
  filter(FDR <= 0.05) %>% 
  filter(logFC < 0.5 ) %>%
  summarise(n_down = n())

# format
dex_comb <- left_join(dex_up, dex_down, by = c("group"))
dex_comb$n_down <- dex_comb$n_down * -1
dex_comb_long <- dex_comb %>% pivot_longer(!group, names_to = "direction", values_to = "n")

# plot
dex_comb_long %>%
  ggplot(aes(x = group, y = n, fill = direction)) + 
  geom_bar(stat="identity", position="identity") +
  geom_text(aes(label=n, vjust = -sign(n))) +
  ggtitle("Dex numbers per group")