Pearson's correlation coefficient with from two tables in r

Question

I have the two data sets below:

df <- read.table(text =
                   "Human_Gene_Name hsapiens    mmusculus   ggallus celegans    dmelanogaster   cintestinalis   trubripes   xtropicalis mmulatta
A1CF    5.634789603 4.787491743 3.688879454 2.079441542 3.931825633 2.772588722 3.871201011 3.044522438 4.094344562
                 AAK1   3.583518938 2.708050201 2.079441542 2.197224577 2.079441542 0.693147181 2.772588722 2.079441542 3.218875825
                 AAMP   3.555348061 3.17805383  2.48490665  1.791759469 2.302585093 0.693147181 2.48490665  1.098612289 2.079441542", header  = T)

ctn_df <- read.table(text = "Species    CTN
                     hsapiens   158
                     mmusculus  85
                     ggallus    67
                     celegans   32
                     dmelanogaster  27
                     cintestinalis  19
                     trubripes  110
                     xtropicalis    82
                     mmulatta   71
                     ", header = T)

The values in 'df' represent functional diveresity, I want to work out the pearsons correlation coefficient for each gene based on the species CTNs and functional diversity.

Is there a way I can easily assign CTN to a specific species in the table 'df' based on the data from 'ctn_df'.

Sorry if this is a simple question.

score 2 · Answer 1 · answered May 09 '18 at 22:41

Use apply to serially pass row numeric values to cor as the first argument and then name the correlation values with the first column:

setNames( apply(df[-1], 1, cor, ctn_df$CTN), df$Human_Gene_Name)
     A1CF      AAK1      AAMP 
0.7556590 0.7834861 0.6829534

score 0 · Answer 2 · answered May 10 '18 at 08:56

Here's a Tidyverse solution:

library(tidyverse)

gather(df, Species, functional_diveresity, -Human_Gene_Name) %>%
  left_join(ctn_df) %>%
  group_by(Human_Gene_Name) %>%
  summarise(cor(functional_diveresity, CTN))

#  # A tibble: 3 x 2
#   Human_Gene_Name `cor(functional_diveresity, CTN)`
#   <fct>                                       <dbl>
# 1 A1CF                                        0.756
# 2 AAK1                                        0.783
# 3 AAMP                                        0.683

The first two lines produce a tidy dataframe which makes downstream calculations easier:

gather(df, Species, functional_diveresity, -Human_Gene_Name) %>%
  left_join(ctn_df)

#    Human_Gene_Name       Species functional_diveresity CTN
# 1             A1CF      hsapiens             5.6347896 158
# 2             AAK1      hsapiens             3.5835189 158
# 3             AAMP      hsapiens             3.5553481 158
# 4             A1CF     mmusculus             4.7874917  85
# 5             AAK1     mmusculus             2.7080502  85
# 6             AAMP     mmusculus             3.1780538  85
# ....

Pearson's correlation coefficient with from two tables in r

2 Answers2