Package 'Elja'

Title: Linear, Logistic and Generalized Linear Models Regressions for the EnvWAS/EWAS Approach
Description: Tool for Environment-Wide Association Studies (EnvWAS / EWAS) which are repeated analysis. It includes three functions. One function for linear regression, a second for logistic regression and a last one for generalized linear models.
Authors: Marwan El Homsi [aut, cre, cph] (ORCID: <https://orcid.org/0000-0002-5782-6757>, affiliation: Desbrest Institute of Epidemiology and Public Health, Univ Montpellier, INSERM, Montpellier, France.), Isabella Annesi-Maesano [ctb, ths, fnd] (ORCID: <https://orcid.org/0000-0002-6340-9300>, affiliation1: Desbrest Institute of Epidemiology and Public Health, Univ Montpellier, INSERM, Montpellier, France., affiliation2: Department of Allergic and Respiratory Diseases, Montpellier University Hospital, Montpellier.)
Maintainer: Marwan El Homsi <[email protected]>
License: GPL (>= 3)
Version: 1.0.1
Built: 2026-09-23 11:48:59 UTC
Source: https://github.com/ehmarwan/elja

Help Index

Generalized Linear Models regression for EnvWAS/EWAS analysis

Description

A tool for Environment-Wide Association Studies (EnvWAS / EWAS) which are repeated analysis. This function is espacially for generalized linear models 'glm' and allows the addition of adjustment variables.

Usage

ELJAglm(
  var,
  var_adjust = NULL,
  family = binomial(link = "logit"),
  data,
  manplot = TRUE,
  nbvalmanplot = 100,
  Bonferroni = FALSE,
  FDR = FALSE,
  manplotsign = FALSE
)

Arguments

var

A categorical and binary variable. It is generally your outcome.
var_adjust

A vector containing the names of the fixed adjustment variables for all the models.
family

The family and the link use for the glm function.
data

A dataframe containing all the variables needed for the analysis.
manplot

Generate a Manhattan plot of the results of the analysis.
nbvalmanplot

The number of variables to include in each Manhattan plot.
Bonferroni

Add a dashed bar to the Manhattan plot showing the Bonferroni significance threshold.
FDR

Add a dashed bar to the Manhattan plot showing the False Discovery Rate (Benjamini-Hochberg method) significance threshold. NA if all p-values > FDR corrected p-values.
manplotsign

Generates a Manhattan plot with only significant results (p<0.05).

Value

A Dataframe with results for each variable of the model.

References

Dunn OJ. Multiple Comparisons Among Means. Journal of the American Statistical Association. 1961;56(293):52‑64. Benjamini Y, Hochberg Y. Controlling the False Discovery Rate: A Practical and Powerful Approach to Multiple Testing. Journal of the Royal Statistical Society: Series B (Methodological). 1995;57(1):289‑300. Leisch F, Dimitriadou E. mlbench: Machine Learning Benchmark Problems. R package version 2.1-11; 2025. Available from: https://CRAN.R-project.org/package=mlbench

Examples

### Loading the synthetic diabetes dataset contained in the mlbench package

library(mlbench)
data(SynthDiabetes)

### Using ELJAlinear to perform EWAS analysis

ELJAglm(var = 'diabetes',data = SynthDiabetes,
family = binomial(link = "logit"), manplot = TRUE, Bonferroni = TRUE,
FDR = TRUE, nbvalmanplot = 30, manplotsign = FALSE)
results

Linear regression for EnvWAS/EWAS analysis

Description

A tool for Environment-Wide Association Studies (EnvWAS / EWAS) namely repeated analyses allowing to estimate the relationships between several environmental factors and a health events. This function is especially for linear regressions and allows the addition of adjustment variables.

Usage

ELJAlinear(
  var,
  var_adjust = NULL,
  data,
  manplot = TRUE,
  nbvalmanplot = 100,
  Bonferroni = FALSE,
  FDR = FALSE,
  manplotsign = FALSE
)

Arguments

var

A categorical and binary variable. It is generally your outcome.
var_adjust

A vector containing the names of the fixed adjustment variables for all the models.
data

A dataframe containing all the variables needed for the analysis.
manplot

Generate a Manhattan plot of the results of the analysis.
nbvalmanplot

The number of variables to include in each Manhattan plot.
Bonferroni

Add a dashed bar to the Manhattan plot showing the Bonferroni significance level.
FDR

Add a dashed bar to the Manhattan plot showing the False Discovery Rate (Benjamini-Hochberg method) significance threshold. NA if all p-values > FDR corrected p-values.
manplotsign

Generates a Manhattan plot with only significant results (p<0.05).

Value

A Dataframe with results for each variable of the model.

References

Dunn OJ. Multiple Comparisons Among Means. Journal of the American Statistical Association. 1961;56(293):52‑64. Benjamini Y, Hochberg Y. Controlling the False Discovery Rate: A Practical and Powerful Approach to Multiple Testing. Journal of the Royal Statistical Society: Series B (Methodological). 1995;57(1):289‑300. Leisch F, Dimitriadou E. mlbench: Machine Learning Benchmark Problems. R package version 2.1-11; 2025. Available from: https://CRAN.R-project.org/package=mlbench

Examples

### Loading the synthetic diabetes dataset contained in the mlbench package

library(mlbench)
data(SynthDiabetes)

### Using ELJAlinear to perform EWAS analysis

ELJAlinear(var = 'pregnant',data = SynthDiabetes,manplot = TRUE,
Bonferroni = TRUE,FDR = TRUE, nbvalmanplot = 30, manplotsign = FALSE)
results

Logistic regression tool for EnvWAS/EWAS analysis

Description

A tool for Environment-Wide Association Studies (EnvWAS / EWAS) which are repeated analysis. This function is espacially for logistic regression based on the glm function with a binomial family with a logit link and allows the addition of adjustment variables.

Usage

ELJAlogistic(
  var,
  var_adjust = NULL,
  data,
  manplot = TRUE,
  nbvalmanplot = 100,
  Bonferroni = FALSE,
  FDR = FALSE,
  manplotsign = FALSE
)

Arguments

var

A categorical and binary variable. It is generally your outcome.
var_adjust

A vector containing the names of the fixed adjustment variables for all the models.
data

A dataframe containing all the variables needed for the analysis.
manplot

Generate a Manhattan plot of the results of the analysis.
nbvalmanplot

The number of variables to include in each Manhattan plot.
Bonferroni

Add a dashed bar to the Manhattan plot showing the Bonferroni significance level.
FDR

Add a dashed bar to the Manhattan plot showing the False Discovery Rate (Benjamini-Hochberg method) significance threshold. NA if all p-values > FDR corrected p-values.
manplotsign

Generates a Manhattan plot with only significant results (p<0.05).

Value

A Dataframe with results for each variable of the model.

References

Dunn OJ. Multiple Comparisons Among Means. Journal of the American Statistical Association. 1961;56(293):52‑64. Benjamini Y, Hochberg Y. Controlling the False Discovery Rate: A Practical and Powerful Approach to Multiple Testing. Journal of the Royal Statistical Society: Series B (Methodological). 1995;57(1):289‑300. Leisch F, Dimitriadou E. mlbench: Machine Learning Benchmark Problems. R package version 2.1-11; 2025. Available from: https://CRAN.R-project.org/package=mlbench

Examples

### Loading the synthetic diabetes dataset contained in the mlbench package

library(mlbench)
data(SynthDiabetes)

### Using ELJAlinear to perform EWAS analysis

ELJAlogistic(var = 'diabetes',data = SynthDiabetes,manplot = TRUE,
Bonferroni = TRUE,FDR = TRUE, nbvalmanplot = 30, manplotsign = FALSE)
results