Divide between groups of rows using group_by

Question

My data is contained in a data.frame:

SYMBOL variable value Sample IDs Group TLR8 MMRF_2613_1_BM 3.186233 Baseline 2613 LessUp TLR8 MMRF_2613_1_BM 5.471014 Baseline 2613 LessUp TLR8 MMRF_2613_1_BM 2.917965 Baseline 2613 MostUp TLR8 MMRF_2613_1_BM 2.147028 Baseline 2613 MostUp TLR4 MMRF_2613_1_BM 7.497424 Baseline 2613 LessUp TLR4 MMRF_2613_1_BM 4.16523 Baseline 2613 LessUp TLR4 MMRF_2613_1_BM 7.136523 Baseline 2613 MostUp TLR4 MMRF_2613_1_BM 7.96523 Baseline 2613 MostUp

For each SYMBOL, I would like to divide the sum of value for the rows where Group is "MostUp" by the sum of value for "LessUp" rows.

I believe I could use the group_by function, but I am not sure how to apply it correctly.

Here is an example of my expected output.

SYMBOL variable value Sample IDs Group TLR8 MMRF_2613_1_BM 0.58 Baseline 2613 MostUp_divided_by_LessUp TLR4 MMRF_2613_1_BM 1.29 Baseline 2613 MostUp_divided_by_LessUp

In addition to calculating the ratios, how would I perform a T-test between the groups?

Can you explain how do you get 0.58 and 1.29 as value ? Do you divide them individually then take mean or sum them and divide? — Ronak Shah
– Ronak Shah, Commented Apr 9, 2020 at 8:49
> (2.147028+2.917965)/(5.471014+3.186233) [1] 0.5850582 > (7.96523+7.136523)/(4.16523+7.497424) [1] 1.294881 — user2300940
– user2300940, Commented Apr 9, 2020 at 8:52

Ronak Shah · Accepted Answer · 2020-04-09 10:06:18Z

We could first calculate the sum of each Group for each Symbol and then divide within each other based on value of 'MostUp' and 'LessUp'.

library(dplyr) df %>% group_by(SYMBOL, variable, Sample, IDs, Group) %>% summarise(value = sum(value)) %>% summarise(value = value[Group == 'MostUp']/value[Group == 'LessUp']) # SYMBOL variable Sample IDs value # <fct> <fct> <fct> <int> <dbl> #1 TLR4 MMRF_2613_1_BM Baseline 2613 1.29 #2 TLR8 MMRF_2613_1_BM Baseline 2613 0.585

To calculate t.test between groups we can do :

df1 <- df %>% group_by(SYMBOL, variable, Sample, IDs) %>% summarise(value = list(t.test(value[Group == 'MostUp'], value[Group == 'LessUp']))) df1 # A tibble: 2 x 5 # Groups: SYMBOL, variable, Sample [2] # SYMBOL variable Sample IDs value # <fct> <fct> <fct> <int> <list> #1 TLR4 MMRF_2613_1_BM Baseline 2613 <htest> #2 TLR8 MMRF_2613_1_BM Baseline 2613 <htest>

data

df <- structure(list(SYMBOL = structure(c(2L, 2L, 2L, 2L, 1L, 1L, 1L, 1L), .Label = c("TLR4", "TLR8"), class = "factor"), variable = structure(c(1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L), .Label = "MMRF_2613_1_BM", class = "factor"), value = c(3.186233, 5.471014, 2.917965, 2.147028, 7.497424, 4.16523, 7.136523, 7.96523), Sample = structure(c(1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L), .Label = "Baseline", class = "factor"), IDs = c(2613L, 2613L, 2613L, 2613L, 2613L, 2613L, 2613L, 2613L), Group = structure(c(1L, 1L, 2L, 2L, 1L, 1L, 2L, 2L ), .Label = c("LessUp", "MostUp"), class = "factor")), class = "data.frame", row.names = c(NA, -8L))

Thanks. What would you do if you should calculate a t-test between the same groups?
@user2300940 you mean something like this? df %>% group_by(SYMBOL, variable, Sample, IDs) %>% summarise(value = list(t.test(value[Group == 'MostUp'], value[Group == 'LessUp'])))
t.test returns output of class 'htest'. Check ?t.test for more details. It is mentioned in the Value section about what values does t.test returns.

Collectives™ on Stack Overflow

Divide between groups of rows using group_by

1 Answer 1

5 Comments

Hot Network Questions

Collectives™ on Stack Overflow

1 Answer 1

5 Comments

Related