Downloading from KEGG

library(volcalc)
library(dplyr) #for left_join()
#> 
#> Attaching package: 'dplyr'
#> The following object is masked from 'package:ChemmineR':
#> 
#>     groups
#> The following objects are masked from 'package:stats':
#> 
#>     filter, lag
#> The following objects are masked from 'package:base':
#> 
#>     intersect, setdiff, setequal, union

The volcalc package can be used to download .mol files directly from KEGG given either compound IDs or pathway IDs.

First, choose a directory to download files to. For this vignette, we will use a temporary directory, but you should choose somewhere in your project.

dl_path <- tempdir()

Single compound usage

You can search KEGG for compunds at https://www.genome.jp/kegg/compound/ to find their KEGG IDs starting with a “C”.

Let’s download .mol files for two compounds, jasmonic acid and methyl jasmonate, with KEGG IDs C08491 and C11512, respectively, using the volcalc function get_mol_kegg().

mols <- get_mol_kegg(compound_ids = c("C08491", "C11512"), dir = dl_path)
mols
#> # A tibble: 2 × 2
#>   compound_id mol_path                                                              
#>   <chr>       <fs::path>                                                            
#> 1 C08491      /var/folders/wr/by_lst2d2fngf67mknmgf4340000gn/T/RtmpKwerlt/C08491.mol
#> 2 C11512      /var/folders/wr/by_lst2d2fngf67mknmgf4340000gn/T/RtmpKwerlt/C11512.mol

The data frame returned by get_mol_kegg() contains the paths the files were downloaded to in mol_path, making for convenient passage on to the volcalc function calc_vol().

rvi <- calc_vol(mols$mol_path)
rvi
#> # A tibble: 2 × 5
#>   mol_path                                                               formula  name                rvi category
#>   <chr>                                                                  <chr>    <chr>             <dbl> <fct>   
#> 1 /var/folders/wr/by_lst2d2fngf67mknmgf4340000gn/T/RtmpKwerlt/C08491.mol C12H18O3 (-)-Jasmonic acid  1.84 moderate
#> 2 /var/folders/wr/by_lst2d2fngf67mknmgf4340000gn/T/RtmpKwerlt/C11512.mol C13H20O3 Methyl jasmonate   3.81 high

calc_vol() also returns the file paths, so these two data frames can be easily joined.

left_join(mols, rvi, by = join_by(mol_path)) %>%
  select(-mol_path)
#> # A tibble: 2 × 5
#>   compound_id formula  name                rvi category
#>   <chr>       <chr>    <chr>             <dbl> <fct>   
#> 1 C08491      C12H18O3 (-)-Jasmonic acid  1.84 moderate
#> 2 C11512      C13H20O3 Methyl jasmonate   3.81 high

Pathway usage

We can download single or multiple compounds with compound_ids, but we can also download all compounds associated with a KEGG pathway with pathway_ids. Let’s download the entire alpha-linolenic acid metabolism pathway (map00592) that the above two compounds are part of.

alam_pathway <- get_mol_kegg(pathway_ids = "map00592", dir = dl_path)
head(alam_pathway)
#> # A tibble: 6 × 3
#>   pathway_id compound_id mol_path                                                                       
#>   <chr>      <chr>       <fs::path>                                                                     
#> 1 map00592   C00157      /var/folders/wr/by_lst2d2fngf67mknmgf4340000gn/T/RtmpKwerlt/map00592/C00157.mol
#> 2 map00592   C01226      /var/folders/wr/by_lst2d2fngf67mknmgf4340000gn/T/RtmpKwerlt/map00592/C01226.mol
#> 3 map00592   C04672      /var/folders/wr/by_lst2d2fngf67mknmgf4340000gn/T/RtmpKwerlt/map00592/C04672.mol
#> 4 map00592   C04780      /var/folders/wr/by_lst2d2fngf67mknmgf4340000gn/T/RtmpKwerlt/map00592/C04780.mol
#> 5 map00592   C04785      /var/folders/wr/by_lst2d2fngf67mknmgf4340000gn/T/RtmpKwerlt/map00592/C04785.mol
#> 6 map00592   C06427      /var/folders/wr/by_lst2d2fngf67mknmgf4340000gn/T/RtmpKwerlt/map00592/C06427.mol
dim(alam_pathway)
#> [1] 44  3

Notice that this returns pathway IDs and compound IDs. We can do the same as above and pass the mol_path column to calc_vol() and then join the resulting data frame and do some basic data wrangling to find the top 10 most volatile compounds in that pathway.

rvi_path <- calc_vol(alam_pathway$mol_path)
#> Warning in FUN(X[[i]], ...): Possible OpenBabel errors detected and only NAs returned.
#> Run with `validate = FALSE` to ignore this.

left_join(alam_pathway, rvi_path, by = join_by(mol_path)) %>% 
  select(-mol_path) %>% 
  #arrange from most to least volatile
  arrange(desc(rvi)) %>% 
  #take just the top 10
  slice_head(n = 10)
#> # A tibble: 10 × 6
#>    pathway_id compound_id formula  name                         rvi category
#>    <chr>      <chr>       <chr>    <chr>                      <dbl> <fct>   
#>  1 map00592   C16310      C6H10O   3-Hexenal                   7.32 high    
#>  2 map00592   C19757      C8H14O2  (3Z)-Hex-3-en-1-yl acetate  6.75 high    
#>  3 map00592   C08492      C6H12O   3-Hexenol                   6.45 high    
#>  4 map00592   C16323      C9H14O   3,6-Nonadienal              6.05 high    
#>  5 map00592   C11512      C13H20O3 Methyl jasmonate            3.81 high    
#>  6 map00592   C16318      C13H20O3 (+)-7-Isomethyljasmonate    3.81 high    
#>  7 map00592   C16322      C9H16O3  9-Oxononanoic acid          2.77 high    
#>  8 map00592   C16343      C17H28O  Heptadecatrienal            2.69 high    
#>  9 map00592   C08491      C12H18O3 (-)-Jasmonic acid           1.84 moderate
#> 10 map00592   C16317      C12H18O3 (+)-7-Isojasmonic acid      1.84 moderate