PomaOutliers: Analyse and Remove Statistical Outliers

In pcastellanoescuder/POMA: Tools for Omics Data Analysis

Analyse and Remove Statistical Outliers

Description

PomaOutliers analyses and removes statistical outliers from the data.

Usage

PomaOutliers(
  data,
  method = "euclidean",
  type = "median",
  outcome = NULL,
  coef = 2,
  labels = FALSE
)

Arguments

data	A SummarizedExperiment object.
method	Character. Indicates the distance measure method to perform MDS.
type	Character. Indicates the type of outlier analysis to perform. Options are "median" (default) and "centroid". See vegan::betadisper.
outcome	Character. Indicates the name of the colData column to be used as the outcome factor. Default is NULL (first factor variable in colData).
coef	Numeric. Indicates the outlier coefficient. Lower values are more sensitive to outliers while higher values are less restrictive about outliers.
labels	Logical. Indicates if sample names should to be plotted.

Value

A list with the results.

Author(s)

Pol Castellano-Escuder

Examples

data <- POMA::st000336 %>% # Example SummarizedExperiment object included in POMA
  PomaImpute() %>% 
  PomaNorm()

## Output is a list with objects `polygon_plot` (ggplot2 object), `distance_boxplot` (ggplot2 object), `outliers` (tibble), and `data` (outlier-free SummarizedExperiment)
outlier_results <- data %>% 
  PomaOutliers(method = "euclidean",
               type = "median",
               outcome = NULL,
               coef = 2,
               labels = FALSE)

outlier_results$data # cleaned SummarizedExperiment object

## Change oulier group factor
outlier_results2 <- data %>% 
  PomaOutliers(method = "euclidean",
               type = "median",
               outcome = "steroids",
               coef = 2,
               labels = FALSE)

outlier_results2$data # cleaned SummarizedExperiment object

pcastellanoescuder/POMA documentation built on Nov. 28, 2024, 1:21 p.m.