Multiple wilcox.tests across columns using variables in first column (R)

Question

I have this data.frame

 df <- data.frame(
      variable=c(2.4860651, -0.68863024, 2.63530974, -2.95754943, 1.67945091, 2.63530974,
           4.79002539, 2.32575938, 3.57236441, -0.364825998, -2.00646016, -3.12380516, 
           0.69307013, -5.65846824, 0.45632519, 2.08978142),
      A=c(0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0),
      B=c(1, 1, 0, 1, 0, 0, 1, 1, 1, 1, 0, 1, 0, 1, 0, 0),
      C=c(0, 0, 1, 0, 0, 0, 0, 1, 1, 0, 0, 0, 1, 0, 1, 1),
      D=c(1, 1, 0, 0, 0, 1, 0, 0, 0, 1, 1, 0, 0, 0, 0, 0),
      E=c(0, 0, 0, 0, 1, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0),
      F=c(0, 1, 0, 1, 0, 1, 0, 0, 0, 0, 0, 0, 1, 1, 0, 1))

I would like to perform wilcox.test for each column with groups defined by 0 and 1 in the columns and using the variables in the column df$variable. Then add the p.values in a new row and adjusted p.values in another row.

I have tried this:

 library(dplyr)
 result <- df %>% summarise(across(!variable, ~wilcox.test(.x ~ variable)$p.value), exact=NULL) %>%
         bind_rows(., p.adjust(., method = 'BH')) %>%
         bind_rows(df, .) %>%
         mutate(variable=replace(variable, is.na(variable), c('p.values', 'p.adjust')))

But this causes errors.

This is the result I would like to get:

 result <- data.frame(
      variable=c(2.4860651, -0.68863024, 2.63530974, -2.95754943, 1.67945091, 2.63530974,
           4.79002539, 2.32575938, 3.57236441, -0.364825998, -2.00646016, -3.12380516, 
           0.69307013, -5.65846824, 0.45632519, 2.08978142, 'p.value', 'p.adjust'),
      A=c(0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 1, 1),
      B=c(1, 1, 0, 1, 0, 0, 1, 1, 1, 1, 0, 1, 0, 1, 0, 0, 0.560444274, 1),
      C=c(0, 0, 1, 0, 0, 0, 0, 1, 1, 0, 0, 0, 1, 0, 1, 1, 0.143117298, 0.764253489),
      D=c(1, 1, 0, 0, 0, 1, 0, 0, 0, 1, 1, 0, 0, 0, 0, 0, 0.820753088, 1),
      E=c(0, 0, 0, 0, 1, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0.95482869, 1),
      F=c(0, 1, 0, 1, 0, 1, 0, 0, 0, 0, 0, 0, 1, 1, 0, 1, 0.254751163, 0.764253489))

Can anyone help?

Ronak Shah · Accepted Answer · 2022-09-29T12:48:03.300

1

You may try something along the lines of the following -

library(dplyr)

tmp <- df %>% summarise(across(!variable, 
          ~wilcox.test(variable[.x == 0], variable[.x == 1])$p.value)) 

adj_value <- p.adjust(unlist(tmp), method = "BH")

result <- bind_rows(df %>% mutate(variable = as.character(variable)), 
          rbind(tmp, adj_value) %>%
                  mutate(variable = c('p.values', 'p.adjust'))
)

edited Sep 29 '22 at 12:48

answered Sep 29 '22 at 02:14

Ronak Shah

377,200
20
156
213

Thank you very much, Ronak. I really like your solution. I just wondered if the result is correct, because for column A zeroes: ```x <- c(2.4860651, -0.68863024, 2.63530974, -2.95754943, 2.63530974, 4.79002539, 2.32575938, 3.57236441, -0.364825998, -2.00646016, -3.12380516, -5.65846824, 0.45632519, 2.08978142)``` and ones: ```y <- c(1.67945091, 0.69307013)``` then ```wilcox.test(x,y)$p.value``` = 1, instead of 0.2789041? Thank you. – Sylvia Rodriguez Sep 29 '22 at 02:44
1

@SylviaRodriguez I think I misunderstood your question earlier. Please see the updated answer. – Ronak Shah Sep 29 '22 at 04:40
Oh this is excellent! Thank you. Just one more question. All ```p.adjust``` are the same as ```p.values```. Do you know how to correct this last issue please? – Sylvia Rodriguez Sep 29 '22 at 04:54
1

It was getting too complex to do it in single pipe so I have broken it down into smaller steps. Hope the update is what you were looking for @SylviaRodriguez – Ronak Shah Sep 29 '22 at 12:48
Thank you Ronak! See also my added answer based on your excellent answers. ¡Gracias! – Sylvia Rodriguez Sep 30 '22 at 08:22

score 0 · Answer 2 · answered Sep 30 '22 at 08:20

Thank you, Ronak. I modified your former answer and I found that this also works and results in the same as you found:

 result <- df %>% 
       summarise(across(!variable, 
           ~wilcox.test(variable[.x == 0], variable[.x == 1])$p.value), exact=NULL) %>%
       bind_rows(., p.adjust(., method = 'BH')) %>%
       bind_rows(df, .) %>% 
       mutate(variable=replace(variable, is.na(variable), c('p.values', 'p.adjust')))

Thank you! :)

Multiple wilcox.tests across columns using variables in first column (R)

2 Answers2