Add stats to the plot

Hello,

I would like to add stats to this plot. Any help would be much appreciated and thanks in advance

Here is my code:

data4.ts%>%  ungroup() %>% t.test(data =.,value ~ Condition) %>%
  adjust_pvalue(method = "bonferroni") 

data2.ts %>%  ungroup() %>% t.test(data =.,value ~ Condition) %>%
  adjust_pvalue(method = "bonferroni") 

a %>% 
  filter(name == "norm_CntTs",
         !is.na(coding)) %>% 
  ggplot(aes(x = Condition, y = value, color = Condition))+
  geom_boxplot(width = 0.5,
               show.legend = FALSE)+
  scale_y_log10()+
  scale_color_manual(
    values = c(
      "HRSV" = "#eb4a40",
      "HRSV_RBV" = "#045275"
    )
  )+
  facet_wrap(.~coding,ncol = 1)
+
  labs(
    title = "Ts",
    y = "Average Ts count" 
  )+
  theme_bw()+
  theme(
    text = element_text(size = 12),
    plot.title = element_text(hjust = 0.5),
    #strip.background = element_rect(color =)
    strip.text = element_text(face = "bold")
  )

enter image description here

• 3,242 views • link

written 2.2 years ago by G.S •

Hi, you can add statistical annotations using ggpubr function stat_pvalue_manual. You need to make sure you calculate the p-values first and then use the stat_pvalue_manual function to add these values to the plot. For example:

data4_test <- data4.ts %>%
  ungroup() %>%
  t.test(value ~ Condition) %>%
  broom::tidy() %>%
  mutate(p.adjusted = p.adjust(p.value, method = "bonferroni"),
         group1 = levels(data4.ts$Condition)[1],
         group2 = levels(data4.ts$Condition)[2])

data2_test <- data2.ts %>%
  ungroup() %>%
  t.test(value ~ Condition) %>%
  broom::tidy() %>%
  mutate(p.adjusted = p.adjust(p.value, method = "bonferroni"),
         group1 = levels(data2.ts$Condition)[1],
         group2 = levels(data2.ts$Condition)[2])

stat_test <- bind_rows(data4_test, data2_test)

a_filtered <- a %>%
  filter(name == "norm_CntTs", !is.na(coding))

plot <- ggplot(a_filtered, aes(x = Condition, y = value, color = Condition)) +
  geom_boxplot(width = 0.5, show.legend = FALSE) +
  scale_y_log10() +
  scale_color_manual(values = c("HRSV" = "#eb4a40", "HRSV_RBV" = "#045275")) +
  facet_wrap(. ~ coding, ncol = 1) +
  labs(title = "Ts", y = "Average Ts count") +
  theme_bw() +
  theme(text = element_text(size = 12),
        plot.title = element_text(hjust = 0.5),
        strip.text = element_text(face = "bold"))

plot_stats <- plot + 
  stat_pvalue_manual(stat_test, 
                     label = "p.adjusted",
                     y.position = max(a_filtered$value) * 1.1, 
                     label.size = 3.5,
                     tip.length = 0.01)

Check the code

• 0 views • link

written 2.2 years ago by marco.barr •

Thanks. That helpful.

I am getting this error

data4_test <- data4.ts%>%  ungroup() %>% t.test(data =.,value ~ Condition)%>%
+   broom::tidy() %>%
+   mutate(p.adjusted = p.adjust(p.value, method = "bonferroni"),
+          group1 = levels(data4.ts$Condition)[1],
+          group2 = levels(data4.ts$Condition)[2])
> 
> data2_test <-data2.ts %>%  ungroup() %>% t
.test(data =.,value ~ Condition) %>%
+   broom::tidy() %>%
+   mutate(p.adjusted = p.adjust(p.value, method = "bonferroni"),
+          group1 = levels(data2.ts$Condition)[1],
+          group2 = levels(data2.ts$Condition)[2])
> 
> stat_test <- bind_rows(data4_test, data2_test)
> 
> a_filtered <- a %>%
+   filter(name == "norm_CntTs", !is.na(coding))
> 
> plot <- ggplot(a_filtered, aes(x = Condition, y = value, color = Condition)) +
+   geom_boxplot(width = 0.5, show.legend = FALSE) +
+   scale_y_log10() +
+   scale_color_manual(values = c("HRSV" = "#eb4a40", "HRSV_RBV" = "#045275")) +
+   facet_wrap(. ~ coding, ncol = 1) +
+   labs(title = "Ts", y = "Average Ts count") +
+   theme_bw() +
+   theme(text = element_text(size = 12),
+         plot.title = element_text(hjust = 0.5),
+         strip.text = element_text(face = "bold"))
> 
> plot_stats <- plot + 
+   stat_pvalue_manual(stat_test, 
+                      label = "p.adjusted",
+                      y.position = max(a_filtered$value) * 1.1, 
+                      label.size = 3.5,
+                      tip.length = 0.01)



Error in asserttat_group_columns_exists(data) : 
  data should contain group1 and group2 columns

• 0 views • link

updated 2.2 years ago by GenoMax

• written 2.2 years ago by G.S •

Please use 101010 to format code so it is represented in monospace font. I have done this for you now.

• 0 views • link

written 2.2 years ago by GenoMax

I think this is what caused the error. we do not have group 1 and 2 in the statistical test results????

enter image description here

• 0 views • link

written 2.2 years ago by G.S •

The error is likely due to the fact that the structure of your data in data4.ts and data2.ts may not contain the Condition values. The group1 and group2 columns are added based on this. The Condition column should be a factor or a categorical variable that distinguishes between the groups you are comparing. Could you show me data4.ts and data2.ts?

• 0 views • link

written 2.2 years ago by marco.barr •

Thanks, thats really helpful

enter image description here

• 0 views • link

written 2.2 years ago by G.S •

Thank you, it seems to be clearer. Your Condition already has two factorial levels. You can replace these names in groups 1 and 2.

data4_test <- data4.ts %>%
  ungroup() %>%
  t.test(value ~ Condition) %>%
  broom::tidy() %>%
  mutate(p.adjusted = p.adjust(p.value, method = "bonferroni"),
         group1 = "HRSV",
         group2 = "HRSV_RBV")

data2_test <- data2.ts %>%
  ungroup() %>%
  t.test(value ~ Condition) %>%
  broom::tidy() %>%
  mutate(p.adjusted = p.adjust(p.value, method = "bonferroni"),
         group1 = "HRSV",
         group2 = "HRSV_RBV")

stat_test <- bind_rows(data4_test, data2_test)

a_filtered <- a %>%
  filter(name == "norm_CntTs", !is.na(coding))

plot <- ggplot(a_filtered, aes(x = Condition, y = value, color = Condition)) +
  geom_boxplot(width = 0.5, show.legend = FALSE) +
  scale_y_log10() +
  scale_color_manual(values = c("HRSV" = "#eb4a40", "HRSV_RBV" = "#045275")) +
  facet_wrap(. ~ coding, ncol = 1) +
  labs(title = "Ts", y = "Average Ts count") +
  theme_bw() +
  theme(text = element_text(size = 12),
        plot.title = element_text(hjust = 0.5),
        strip.text = element_text(face = "bold"))

plot_stats <- plot + 
  stat_pvalue_manual(stat_test, 
                     label = "p.adjusted",
                     y.position = max(a_filtered$value) * 1.1, 
                     label.size = 3.5,
                     tip.length = 0.01)

Try and let me know

• 0 views • link

written 2.2 years ago by marco.barr •

if don't work try binding the stats without group.

stat_data4 <- data4.ts %>%
  ungroup() %>%
  t.test(data = ., value ~ Condition) %>%
  broom::tidy() %>%
  mutate(p.adj = p.adjust(p.value, method = "bonferroni"))

stat_data2 <- data2.ts %>%
  ungroup() %>%
  t.test(data = ., value ~ Condition) %>%
  broom::tidy() %>%
  mutate(p.adj = p.adjust(p.value, method = "bonferroni"))
stat_data <- bind_rows(
  stat_data4 %>% mutate(dataset = "data4"),
  stat_data2 %>% mutate(dataset = "data2")
)

your plot + stat_pvalue_manual(stat_data, label = "p.adj", tip.length = 0.01)

the other parts of the code remain the same. I hope this work, let me know.

• 0 views • link

written 2.2 years ago by marco.barr •

Thanks for the help. Unfortunately, did not work :(

error in asserttat_group_columns_exists(data) : data should contain group1 and group2 columns

• 0 views • link

written 2.2 years ago by G.S •

1 answer

Finally, It works. I have added this code

my_comparisons=list(c("HRSV", "HRSV_RBV"))

plot + stat_compare_means(method = "wilcox.test",comparisons = my_comparisons, label.y = 4.5, size= 4, hjust= 0.09,vjust=0.7, alternative = "two.sided",
                          paired = F, exact = F, label = "p.value")

enter image description here

• 0 views • link

written 2.2 years ago by G.S •