Help for package phoenics

Type:

Package

Title:

Pathways Longitudinal and Differential Analysis in Metabolomics

Version:

0.5

Date:

2025-01-29

Depends:

R (≥ 4.0.0)

Imports:

lme4, blme, tibble, FactoMineR, factoextra, tidyr, stats

Suggests:

knitr, KEGGREST

Maintainer:

Camille Guilmineau <camille.guilmineau@inrae.fr>

Description:

Perform a differential analysis at pathway level based on metabolite quantifications and information on pathway metabolite composition. The method is based on a Principal Component Analysis step and on a linear mixed model. Automatic query of metabolic pathways is also implemented.

License:

GPL-3

Repository:

CRAN

BugReports:

https://forgemia.inra.fr/panoramics/phoenics/-/issues

Encoding:

UTF-8

NeedsCompilation:

URL:

https://forgemia.inra.fr/panoramics/phoenics

LazyData:

true

RoxygenNote:

7.3.2

VignetteBuilder:

knitr

Language:

en-US

Packaged:

2025-01-29 12:57:25 UTC; nathalie

Author:

Camille Guilmineau [aut, cre], Remi Servien

[aut], Nathalie Vialaneix

[aut]

Date/Publication:

2025-01-29 14:20:06 UTC

Dataset "MTBLS422"

Description

Metabolites quantifications, associated metabolic pathways and experimental design to characterize the effects of two clinically important antibiotic treatments, ciprofloxacin and vancomycin-imipenem on mice.

Format

3 datasets are provided:

quantif: a data frame with 10 rows (metabolites name) and 11 columns (samples id), which corresponds to the metabolites quantifications in the samples.
pathways: a data frame with 11 rows and 4 columns, which contains informations about pathways and their metabolites.
design: a data frame with 11 rows (samples id) and 4 columns (id and effects to be added in the model).

Details

The raw dataset have been made available on MetaboLights (with the id MTBLS422 https://www.ebi.ac.uk/metabolights/editor/MTBLS422) by [Choo et al., 2017]. Metabolite quantifications were obtained based on the raw signal using ASICS package. Pathways were obtained using KEGGREST package. The datasets provided for the example are a subset of the original dataset.

References

Choo J. M., Kanno T., Zain N. M. M., Leong L. E. X., Abell G. C. J., Keeble J. E., Bruce K. D., Mason A. J., Rogers G. B. (2017). Divergent relationships between fecal microbiota and metabolome following distinct antibiotic-induced disruptions. mSphere, 2(1) doi:10.1128/msphere.00005-17

Lefort G., Liaubet L., Canlet C., Tardivel P., P\'ere M.C., Quesnel H., Paris A., Iannuccelli N., Vialaneix N. Servien R. (2019). ASICS: an R package for a whole analysis workflow of 1D 1H NMR spectra. Bioinformatics, 35(21): 4356–4363. doi:10.1093/bioinformatics/btz248

Tardivel P., Canlet C., Lefort G., Tremblay-Franco M., Debrauwer L., Concordet D., Servien R. (2017). ASICS: an automatic method for identification and quantification of metabolites in complex 1D 1H NMR spectra. Metabolomics, 13(10): 109. doi:10.1007/s11306-017-1244-5

Tenenbaum D., Maintainer B. (2022). KEGGREST: Client-side REST access to the Kyoto Encyclopedia of Genes and Genomes (KEGG). R package version 1.38.0.

Examples

data(MTBLS422)

design[1:5, ]
pathways[1:5, ]
quantif[1:5, 1:5]

Prepare quantification data from ASICS outputs

Description

Prepare quantification data from ASICS outputs for test_pathway. In short, it replaces metabolite names by metabolites KEGG codes and transposes the matrix to have samples in rows and metabolites in columns.

Usage

from_ASICS_to_PHOENICS(quantif)

Arguments

quantif

output matrix of ASICS quantification

Value

A matrix of quantification with samples in rows and metabolites in columns, properly formatted for test_pathway

Author(s)

Camille Guilmineau <camille.guilmineau@inrae.fr>
Remi Servien <remi.servien@inrae.fr>
Nathalie Vialaneix <nathalie.vialaneix@inrae.fr>

References

Examples

data("MTBLS422")
quantif <- from_ASICS_to_PHOENICS(quantif)

Calculate overlap coefficient between pathways

Description

Calculate overlap coefficient between pathways

Usage

overlap_coefficient(pathwayA, pathwayB, pathways = NULL, organism = NULL)

Arguments

pathwayA

a character string of pathway name or pathway code or a character vector of metabolite names or metabolite codes

pathwayB

a character string of pathway name or pathway code or a character vector of metabolite names or metabolite codes

pathways

data.frame or matrix with metabolites in rows and the following information in columns:

metabolite_code metabolite code
metabolite_name metabolite name
pathway_code pathway code (identifier)
pathway_name name of the pathway

Used if pathwayA and pathwayB are pathway names or pathway codes.

organism

organism code in KEGG database. Required if pathways = NULL and pathwayA and pathwayB are pathway names and ignored otherwise.

Value

A value between 0 and 1 calculated with the formula:

\text{OC}(A,B) = \frac{\vert A \cap B \vert}{\min(\vert A \vert, \vert B \vert)}

An overlap coefficient of 1 means that one pathway is included in the other. An overlap coefficient of 0 means that there is no overlap between the pathways.

References

Wieder C., Lai R.P.J., Ebbels T.M.D. (2022). Single sample pathway analysis in metabolomics: performance evaluation and application. BMC Bioinformatics, 23, 481. doi:10.1186/s12859-022-05005-1

Examples

data("MTBLS422")
pathwayA <- "Galactose metabolism"
pathwayB <- "Vitamin digestion and absorption"
overlap_coefficient(pathwayA, pathwayB, pathways)

if (requireNamespace("KEGGREST", quietly = TRUE)) {
  pathwayA <- "Galactose metabolism"
  pathwayB <- "Vitamin digestion and absorption"
  
  overlap_coefficient(pathwayA, pathwayB, organism = "mmu")
  
}

pathwayA <- "mmu00052"
pathwayB <- "mmu00562"
overlap_coefficient(pathwayA, pathwayB, pathways)

pathwayA <- c("C00029", "C00116", "C00137", "C00794", "C00984", "C01697")
pathwayB <- c("C00191", "C00092", "C00137")
overlap_coefficient(pathwayA, pathwayB)

Class pathwayRes

Description

S3 class for pathway differential analysis results

Usage

## S3 method for class 'pathwayRes'
summary(object, ...)

## S3 method for class 'pathwayRes'
print(x, ...)

## S3 method for class 'pathwayRes'
plot(
  x,
  ...,
  pathway_id = NULL,
  plot = c("eig", "var", "ind", "group"),
  habillage = "none"
)

## S3 method for class 'pathwayRes'
head(x, ...)

extract(object, pathway_id)

adjust_pval(object, method = p.adjust.methods, n = length(object))

Arguments

object, x

object of class pathwayRes

...

not used

pathway_id

a character string or vector of pathway codes or names

plot

a character string indicating the type of plot to return. Default to "eig" (the screegraph of the PCA or MFA is displayed)

habillage

a character string indicating the column of the design used to color the individuals. Only used when plot = "ind". Default to "none" (no color)

method

a character string indicating the correction method to be used for multiple testing correction (authorized values are those of p.adjust.methods)

n

number of comparisons for multiple testing correction

Details

Methods for the class pathwayRes

Value

The function extract returns an object of class pathwayRes which is a list of pathway results, containing only the pathways in pathway_id.

The function adjust_pval returns a data.frame with pathways in rows and the following information in columns:

pathway_name

name of the pathway

pathway_code

pathway code (identifier)

Fixed_effect

tested effect

pval

raw p-value of the pathway

adjusted_pval

adjusted p-value of the pathway

Examples

data("MTBLS422")
quantif <- from_ASICS_to_PHOENICS(quantif)
out_test <- test_pathway(quantif, design, pathways, 
                         fixed = c("Age", "Treatment"), random = "Mouse", 
                         npc = 2, model = "blmer")
summary(out_test)
out_test2 <- test_pathway(quantif, design, pathways, 
                         fixed = c("Age", "Treatment"), random = "Mouse", 
                         npc = 2, model = "blmer", analysis = "MFA")
summary(out_test2)
print(out_test)
print(out_test2)
plot(out_test)
plot(out_test, pathway_id = "mmu00052", plot = "var")
plot(out_test, pathway_id = "mmu00052", plot = "ind", habillage = "Age")
plot(out_test2, pathway_id = "mmu00562", plot = "eig")
plot(out_test2, pathway_id = "mmu00562", plot = "var")
plot(out_test2, pathway_id = "mmu00562", plot = "ind")
plot(out_test2, pathway_id = "mmu00562", plot = "group")
extract(out_test, "mmu00562")
adj_pval <- adjust_pval(out_test)

Query KEGG pathways for a given set of metabolites

Description

Query KEGG pathways for a given set of metabolites

Usage

pathway_search(metab, organism, min_size = 2)

Arguments

metab

vector of metabolite KEGG codes

organism

organism code in KEGG database

min_size

minimal number of metabolites required for a pathway to be returned

Value

a data.frame with metabolites in rows and the following information in columns:

metabolite_code metabolite code
metabolite_name metabolite name
pathway_code pathway code (identifier)
pathway_name name of the pathway

Author(s)

Camille Guilmineau <camille.guilmineau@inrae.fr>
Remi Servien <remi.servien@inrae.fr>
Nathalie Vialaneix <nathalie.vialaneix@inrae.fr>

References

Kanehisa M., Goto S. (2000). KEGG: Kyoto Encyclopedia of Genes and Genomes, Nucleic Acids Research, Volume 28, Issue 1, Pages 27–30, doi:10.1093/nar/28.1.27

Tenenbaum D., Maintainer B. (2022). KEGGREST: Client-side REST access to the Kyoto Encyclopedia of Genes and Genomes (KEGG). R package version 1.38.0.

Examples

if (requireNamespace("KEGGREST", quietly = TRUE)) {
  data("MTBLS422")
  quantif <- from_ASICS_to_PHOENICS(quantif)
  
  pathways <- pathway_search(metab = colnames(quantif), organism = "mmu")
  
}

Pathway differential analysis based on longitudinal metabolomics data

Description

Perform a differential analysis at pathway level based on metabolite quantifications and information on pathway metabolite composition. The method relies on a PCA or a MFA step.

Usage

test_pathway(
  quantif,
  design,
  pathways = "auto",
  fixed,
  random,
  npc = 1L,
  model = c("lmer", "blmer"),
  organism = NULL,
  min_size = 2,
  analysis = c("PCA", "MFA")
)

Arguments

quantif

data.frame or matrix of the metabolite quantification with samples in rows (sample identifiers must be row names) and metabolites in columns (metabolite codes must be column names)

design

data.frame or matrix with samples in rows (sample identifiers must be row names) and the different effects to be included in the model in columns. Column names must be provided and are used as arguments for fixed and random

pathways

data.frame or matrix with metabolites in rows (all metabolites in columns of quantif must have a row in this input) and the following information in columns:

metabolite_code metabolite code
metabolite_name metabolite name
pathway_code pathway code (identifier)
pathway_name name of the pathway

fixed

character vector of fixed effects to be included in the model. They must correspond to column names of design. If analysis = "MFA", the first fixed effect must correspond to the time effect

random

character vector of random effects to be included in the model. They must correspond to column names of design

npc

number of PCs for the PCA step

model

a character string indicating if the model has to be fitted using lmer or blmer. Default to "lmer"

organism

organism code in KEGG database. Required if pathways = "auto" and ignored otherwise

min_size

minimal number of metabolites in a pathway. Required if pathways = "auto" and ignored otherwise. Default to 2

analysis

character string indicating if the pathway scores are obtained using PCA or MFA

Details

If pathways = "auto", information on pathways in which metabolites are present is automatically obtained by the function pathway_search using KEGGREST that queries KEGG database. Results are specific to a given organism (passed in organism). Pathways containing less than min_size metabolites are filtered out.

Value

an object of class PCApath or MFApath that inherits from class pathwayRes (a list of pathway results). Each element of the list contains the following entries:

pathway_name

a character corresponding to the pathway name

pathway_code

a character corresponding to the pathway code

metabolites

a data.frame with the names and codes of the quantified metabolites in the pathway

PCA

the result of the pathway PCA or MFA (a PCA object as obtained from PCA or a MFA object as obtained from MFA)

model

the output of the mixed model fit

test_pathway

a data.frame with the p-values for each tested fixed effect

Author(s)

Camille Guilmineau <camille.guilmineau@inrae.fr>
Remi Servien <remi.servien@inrae.fr>
Nathalie Vialaneix <nathalie.vialaneix@inrae.fr>

Examples

data("MTBLS422")
quantif <- from_ASICS_to_PHOENICS(quantif)
out_test <- test_pathway(quantif, design, pathways, 
                         fixed = c("Age", "Treatment"), random = "Mouse", 
                         npc = 2, model = "blmer")
out_test

out_test2 <- test_pathway(quantif, design, pathways, 
                         fixed = c("Age", "Treatment"), random = "Mouse", 
                         npc = 2, model = "blmer", analysis = "MFA")
out_test2

if (requireNamespace("KEGGREST", quietly = TRUE)) { 

  out_test3 <- test_pathway(quantif, design, pathways = "auto", 
                            fixed = c("Age", "Treatment"), random = "Mouse", 
                            npc = 2, model = "blmer", organism = "mmu")
  out_test3
 
}

Dataset "MTBLS422"

Description

Format

Details

References

Examples

Prepare quantification data from ASICS outputs

Description

Usage

Arguments

Value

Author(s)

References

Examples

Calculate overlap coefficient between pathways

Description

Usage

Arguments

Value

References

Examples

Class pathwayRes

Description

Usage

Arguments

Details

Value

Examples

Query KEGG pathways for a given set of metabolites

Description

Usage

Arguments

Value

Author(s)

References

Examples

Pathway differential analysis based on longitudinal metabolomics data

Description

Usage

Arguments

Details

Value

Author(s)

See Also

Examples