The retention time is a crucial piece of information for metabolite identification. However, the retention time of the same metabolite can vary under different chromatographic conditions and parameters, leading to discrepancies between experimental and library retention times. MetaboRI is a retention index-based program designed to address these issues.
The R package can be installed and used through the following methods:
devtools::install_github("dongdongdong7/MetaboRI")
library(MetaboRI)Endogenous retention time correction metabolite have three types: Carnitine, Dicarboxylic acid, and Hydroxy fatty acid. These metabolites have several advantages: (1) wide coverage of retention time; (2) endogenous biological metabolites; (3) retention time is related to chain length, making it easier to predict.
The following format (lc_tibble) is used to store endogenous retention time-corrected metabolites:
> data("standard_car", )
> standard_car
# A tibble: 16 × 11
id name rt lc ri exactmass formula smiles inchi inchikey predicted
<chr> <chr> <dbl> <dbl> <dbl> <dbl> <chr> <chr> <chr> <chr> <lgl>
1 HMDB0000062 L-Carnitine 0.52 0 0 162. C7H15NO3 C[N+](C)(C)C[C@H]… InCh… PHIQHXF… FALSE
2 HMDB0000201 L-Acetylcarnitine 0.92 2 200 204. C9H17NO4 CC(=O)O[C@H](CC(O… InCh… RDHQFKQ… FALSE
3 HMDB0000824 Propionylcarnitine 2.24 3 300 218. C10H19NO4 CCC(=O)O[C@H](CC(… InCh… UFAHZIU… FALSE
4 HMDB0002013 Butyrylcarnitine 3.08 4 400 231. C11H21NO4 CCCC(=O)O[C@H](CC… InCh… QWYFHHG… FALSE
5 HMDB0013128 Valerylcarnitine 3.92 5 500 246. C12H23NO4 CCCCC(=O)O[C@H](C… InCh… VSNFQQX… FALSE
6 HMDB0000756 Hexanoylcarnitine 4.56 6 600 260. C13H25NO4 CCCCCC(=O)O[C@H](… InCh… VVPRQWT… FALSE
7 HMDB0000791 Octanoylcarnitine 5.52 8 800 288. C15H29NO4 CCCCCCCC(=O)O[C@H… InCh… CXTATJF… FALSE
8 HMDB0000651 Decanoylcarnitine 6.29 10 1000 315. C17H33NO4 CCCCCCCCCC(=O)O[C… InCh… LZOSYCM… FALSE
9 HMDB0002250 Dodecanoylcarnitine 6.91 12 1200 344. C19H37NO4 CCCCCCCCCCCC(=O)O… InCh… FUJLYHJ… FALSE
10 HMDB0005066 Tetradecanoylcarnitine 7.48 14 1400 372. C21H41NO4 CCCCCCCCCCCCCC(=O… InCh… PSHXNVG… FALSE
11 HMDB0000222 Palmitoylcarnitine 7.97 16 1600 400. C23H45NO4 CCCCCCCCCCCCCCCC(… InCh… XOMRRQX… TRUE
12 HMDB0000848 Stearoylcarnitine 8.42 18 1800 428. C25H49NO4 CCCCCCCCCCCCCCCCC… InCh… FNPHNLN… FALSE
13 HMDB0006460 Arachidyl carnitine 8.8 20 2000 455. C27H53NO4 CCCCCCCCCCCCCCCCC… InCh… SVJLJQB… FALSE
14 HMDB0062468 Docosanoylcarnitine 9.12 22 2200 483. C29H57NO4 CCCCCCCCCCCCCCCCC… InCh… IUMXSSO… FALSE
15 HMDB0240665 Lignoceroylcarnitine 9.58 24 2400 511. C31H61NO4 CCCCCCCCCCCCCCCCC… InCh… YDUFZFU… FALSE
16 HMDB0006347 Hexacosanoyl carnitine 9.68 26 2600 539. C33H65NO4 CCCCCCCCCCCCCCCCC… InCh… KOCKWDD… TRUE > data("standard_dic")
> standard_dic
# A tibble: 25 × 11
id name rt lc ri exactmass formula smiles inchi inchikey predicted
<chr> <chr> <dbl> <dbl> <dbl> <dbl> <chr> <chr> <chr> <chr> <lgl>
1 HMDB0002329 Oxalic acid 0.54 2 200 90.0 C2H2O4 OC(=O)C(O)=O InChI=… MUBZPKH… FALSE
2 HMDB0000691 Malonic acid 0.73 3 300 104. C3H4O4 OC(=O)CC(O)=O InChI=… OFOBLEO… FALSE
3 HMDB0000254 Succinic acid 1.4 4 400 118. C4H6O4 OC(=O)CCC(O)=O InChI=… KDYFGRW… FALSE
4 HMDB0000661 Glutaric acid 2.39 5 500 132. C5H8O4 OC(=O)CCCC(O)=O InChI=… JFCQEDH… FALSE
5 HMDB0000448 Adipic acid 3.15 6 600 146. C6H10O4 OC(=O)CCCCC(O)=O InChI=… WNLRTRB… FALSE
6 HMDB0000857 Pimelic acid 3.91 7 700 160. C7H12O4 OC(=O)CCCCCC(O)=O InChI=… WLJVNTC… FALSE
7 HMDB0000893 Suberic acid 4.52 8 800 174. C8H14O4 OC(=O)CCCCCCC(O)=O InChI=… TYFQFVW… FALSE
8 HMDB0000784 Azelaic acid 5.05 9 900 188. C9H16O4 OC(=O)CCCCCCCC(O)=O InChI=… BDJRBEY… FALSE
9 HMDB0000792 Sebacic acid 5.53 10 1000 202. C10H18O4 OC(=O)CCCCCCCCC(O)=O InChI=… CXMXRPH… FALSE
10 HMDB0000888 Undecanedioic acid 5.98 11 1100 216. C11H20O4 OC(=O)CCCCCCCCCC(O)=O InChI=… LWBHHRR… FALSE
# ℹ 15 more rows
# ℹ Use `print(n = ...)` to see more rows> data("standard_hyd")
> standard_hyd
# A tibble: 21 × 11
id name rt lc ri exactmass formula smiles inchi inchikey predicted
<chr> <chr> <dbl> <dbl> <dbl> <dbl> <chr> <chr> <chr> <chr> <lgl>
1 HMDB0000115 Glycolic acid 0.63 2 200 76.0 C2H4O3 OCC(O)=O InCh… AEMRFAO… FALSE
2 HMDB0000190 L-Lactic acid 0.9 3 300 90.0 C3H6O3 C[C@H](O)C(O)=O InCh… JVTAAEK… FALSE
3 CID11266 2-Hydroxybutyric acid 1.82 4 400 104. C4H8O3 CCC(C(=O)O)O InCh… AFENDNX… FALSE
4 HMDB0001863 2-Hydroxyvaleric acid 3.19 5 500 118. C5H10O3 CCCC(O)C(O)=O InCh… JRHWHSJ… FALSE
5 HMDB0001624 2-Hydroxycaproic acid 4.32 6 600 132. C6H12O3 CCCCC(O)C(O)=O InCh… NYHNVHG… FALSE
6 CID2750949 2-hydroxyheptanoic acid 5.16 7 700 146. C7H14O3 CCCCCC(C(=O)O)O InCh… RGMMREB… FALSE
7 HMDB0000711 2-Hydroxyoctanoic acid 5.83 8 800 160. C8H16O3 CCCCCCC(O)C(O)=O InCh… JKRDADV… FALSE
8 CID5282897 2-Hydroxynonanoic acid 6.43 9 900 174. C9H18O3 CCCCCCCC(C(=O)O)O InCh… BTJFTHO… FALSE
9 HMDB0242148 2-Hydroxycapric acid 6.96 10 1000 188. C10H20O3 CCCCCCCCC(O)C(O)… InCh… GHPVDCP… FALSE
10 HMDB0059736 2-hydroxyundecanoic acid 7.43 11 1100 202. C11H22O3 CCCCCCCCCC(O)C(O… InCh… MNRBGFK… FALSE
# ℹ 11 more rows
# ℹ Use `print(n = ...)` to see more rowsMetaboRI requires the user to prepare the following tibble as input. The correction metabolite carnitine is used as an example:
Experimental data, including feature id and corresponding retention time.
> data("experimental_data")
> experimental_data
# A tibble: 4,629 × 2
feature rt
<chr> <dbl>
1 0.05_97.9687m/z 0.0501
2 0.14_136.0734m/z 0.142
3 0.21_391.8773n 0.214
4 0.43_308.2184m/z 0.431
5 0.45_234.9509m/z 0.451
6 0.45_188.0227m/z 0.451
7 0.45_161.0119m/z 0.451
8 0.45_200.9741m/z 0.451
9 0.45_170.0658m/z 0.451
10 0.45_146.1651m/z 0.451
# ℹ 4,619 more rows
# ℹ Use `print(n = ...)` to see more rowsThe retention time correction metabolite under experimental chromatographic conditions, using carnitine as an example. The retention time correction metabolite with unknown retention time can be set to NA and the predicted value set to TRUE.
> data("experimental_car")
> experimental_car
# A tibble: 16 × 11
id name rt lc ri exactmass formula smiles inchi inchikey predicted
<chr> <chr> <dbl> <int> <dbl> <dbl> <chr> <chr> <chr> <chr> <lgl>
1 HMDB0000062 L-Carnitine 0.55 0 0 162. C7H15NO3 C[N+](C)(C)C[C@H]… InCh… PHIQHXF… FALSE
2 HMDB0000201 L-Acetylcarnitine 0.91 2 200 204. C9H17NO4 CC(=O)O[C@H](CC(O… InCh… RDHQFKQ… FALSE
3 HMDB0000824 Propionylcarnitine 2.03 3 300 218. C10H19NO4 CCC(=O)O[C@H](CC(… InCh… UFAHZIU… FALSE
4 HMDB0002013 Butyrylcarnitine 3.03 4 400 231. C11H21NO4 CCCC(=O)O[C@H](CC… InCh… QWYFHHG… FALSE
5 HMDB0013128 Valerylcarnitine 3.9 5 500 246. C12H23NO4 CCCCC(=O)O[C@H](C… InCh… VSNFQQX… FALSE
6 HMDB0000756 Hexanoylcarnitine 4.63 6 600 260. C13H25NO4 CCCCCC(=O)O[C@H](… InCh… VVPRQWT… FALSE
7 HMDB0000791 Octanoylcarnitine 5.65 8 800 288. C15H29NO4 CCCCCCCC(=O)O[C@H… InCh… CXTATJF… FALSE
8 HMDB0000651 Decanoylcarnitine 6.48 10 1000 315. C17H33NO4 CCCCCCCCCC(=O)O[C… InCh… LZOSYCM… FALSE
9 HMDB0002250 Dodecanoylcarnitine 7.17 12 1200 344. C19H37NO4 CCCCCCCCCCCC(=O)O… InCh… FUJLYHJ… FALSE
10 HMDB0005066 Tetradecanoylcarnitine 7.76 14 1400 372. C21H41NO4 CCCCCCCCCCCCCC(=O… InCh… PSHXNVG… FALSE
11 HMDB0000222 Palmitoylcarnitine 8.28 16 1600 400. C23H45NO4 CCCCCCCCCCCCCCCC(… InCh… XOMRRQX… FALSE
12 HMDB0000848 Stearoylcarnitine 8.74 18 1800 428. C25H49NO4 CCCCCCCCCCCCCCCCC… InCh… FNPHNLN… FALSE
13 HMDB0006460 Arachidyl carnitine NA 20 2000 455. C27H53NO4 CCCCCCCCCCCCCCCCC… InCh… SVJLJQB… TRUE
14 HMDB0062468 Docosanoylcarnitine NA 22 2200 483. C29H57NO4 CCCCCCCCCCCCCCCCC… InCh… IUMXSSO… TRUE
15 HMDB0240665 Lignoceroylcarnitine NA 24 2400 511. C31H61NO4 CCCCCCCCCCCCCCCCC… InCh… YDUFZFU… TRUE
16 HMDB0006347 Hexacosanoyl carnitine NA 26 2600 539. C33H65NO4 CCCCCCCCCCCCCCCCC… InCh… KOCKWDD… TRUEdata("experimental_data")
data("experimental_car")
data("standard_car")ResList <- lc2rt(lc_tibble = experimental_car, targetRt = max(experimental_data$rt))
experimental_car <- ResList$lc_tibble
ResList$p
ResList <- lc2rt(lc_tibble = standard_car, targetRt = max(experimental_data$rt))
standard_car <- ResList$lc_tibble
ResList$p
It is recommended that a lower limit be set for the range of retention time corrections, which is typically the value with the minimum retention time of experimental metabolite feature or the dead time.
experimental_car <- deadTime_add(lc_tibble = experimental_car, deadTime = min(experimental_data$rt))
standard_car <- deadTime_add(lc_tibble = standard_car, deadTime = min(experimental_data$rt))experimental_data$ri <- sapply(experimental_data$rt, function(x) {
rt2ri(x, experimental_lc = experimental_car)
})> experimental_data
# A tibble: 4,629 × 3
feature rt ri
<chr> <dbl> <dbl>
1 0.05_97.9687m/z 0.0501 -100
2 0.14_136.0734m/z 0.142 -81.7
3 0.21_391.8773n 0.214 -67.2
4 0.43_308.2184m/z 0.431 -23.7
5 0.45_234.9509m/z 0.451 -19.9
6 0.45_188.0227m/z 0.451 -19.9
7 0.45_161.0119m/z 0.451 -19.9
8 0.45_200.9741m/z 0.451 -19.9
9 0.45_170.0658m/z 0.451 -19.9
10 0.45_146.1651m/z 0.451 -19.9
# ℹ 4,619 more rows
# ℹ Use `print(n = ...)` to see more rows# Make the experimental lc_tibble and standard lc_tibble have the same number of lc.
ResList <- compare_lc_tibble(experimental_lc = experimental_car, standard_lc = standard_car)
experimental_car <- ResList$experimental_lc
standard_car <- ResList$standard_lc
# Calculate new retention time
experimental_data <- orignRT2newRT(experimental_data = experimental_data,
experimental_lc = experimental_car,
standard_lc = standard_car, thread = 1)> experimental_data
# A tibble: 4,629 × 4
feature orign_rt ri new_rt
<chr> <dbl> <dbl> <dbl>
1 0.05_97.9687m/z 0.0501 -100 0.0501
2 0.14_136.0734m/z 0.142 -81.7 0.136
3 0.21_391.8773n 0.214 -67.2 0.204
4 0.43_308.2184m/z 0.431 -23.7 0.409
5 0.45_234.9509m/z 0.451 -19.9 0.427
6 0.45_188.0227m/z 0.451 -19.9 0.427
7 0.45_161.0119m/z 0.451 -19.9 0.427
8 0.45_200.9741m/z 0.451 -19.9 0.427
9 0.45_170.0658m/z 0.451 -19.9 0.427
10 0.45_146.1651m/z 0.451 -19.9 0.427
# ℹ 4,619 more rows
# ℹ Use `print(n = ...)` to see more rows