PomaLMM: Linear Mixed Models

In pcastellanoescuder/POMA_package: Tools for Omics Data Analysis

Linear Mixed Models

Description

PomaLMM performs linear mixed models on a SummarizedExperiment object.

Usage

PomaLMM(data, x = NULL, y = NULL, adjust = "fdr", clean_plot = FALSE)

Arguments

data	A SummarizedExperiment object.
x	Character vector. Indicates the names of colData columns to be used as random and fixed effects (independent variables). If it's set to NULL (default), all variables in colData will be used.
y	Character vector. Indicates the names of dependent variables. If it's NULL (default), all features will be used.
adjust	Character. Multiple comparisons correction method to adjust p-values. Available options are: "fdr" (false discovery rate), "holm", "hochberg", "hommel", "bonferroni", "BH" (Benjamini-Hochberg), and "BY" (Benjamini-Yekutieli).
clean_plot	Logical. Indicates if remove intercept and linear mixed model residues boxplots from the plot. Defasult is FALSE.

Value

A list with results including plots and tables. Table values indicate the percentage variance explained per variable.

Author(s)

Pol Castellano-Escuder

Examples

data <- POMA::st000284 %>% # Example SummarizedExperiment object included in POMA
  PomaImpute() %>% 
  PomaNorm()

## Output is a list with objects `lm_table` (tibble) and `regression_plot` (ggplot2 object)
## Perform linear mixed model with all features
#data %>%
#  PomaLMM()
#
## Perform linear mixed model with two features
#data %>% 
#  PomaLMM(y = c("x1_methyladenosine", "x1_methylhistamine"))
#
## Perform linear mixed model with one random effect
#data %>% 
#  PomaLMM(x = "smoking_condition")
#
## Perform linear mixed model with two random effects and two features
#data %>% 
#  PomaLMM(x = c("smoking_condition", "gender"),
#          y = c("x1_methyladenosine", "x1_methylhistamine"))
#
## Perform linear mixed model with no random effects and two features, therefore, a linear model will be fitted
#data %>% 
#  PomaLMM(x = "age_at_consent", # Numerical, i.e., fixed effect
#          y = c("x1_methyladenosine", "x1_methylhistamine"))
#
## Perform linear mixed model with no random effects and all features, therefore, a linear model will be fitted
#data %>% 
#  PomaLMM(x = "age_at_consent") # Numerical i.e., fixed effect

pcastellanoescuder/POMA_package documentation built on Nov. 28, 2024, 1:23 p.m.