MWAS_filter: Filter MWAS results by p-value and/or CV

In MWASTools: MWASTools: an integrated pipeline to perform metabolome-wide association studies

Description

This function allows filtering the output matrix from "MWAS_stats()", by p-value and/or coefficient of variation (CV).

Usage

MWAS_filter(MWAS_matrix, type = "pvalue", alpha_th = 0.05, CV_th = 0.30, sort = FALSE)

Arguments

MWAS_matrix	numeric matrix generated by the function "MWAS_stats()".
type	character constant indicating the filtering criteria. If type = "pvalue", only metabolic variables with p-value below alpha_th will be retained in the MWAS_matrix. If type = "CV", only metabolic variables with CV below CV_th will be retained. If type = "all", only metabolic variables with CV below CV_th and p-value below alpha_th will be retained.
alpha_th	numeric value indicating the significance threshold.
CV_th	numeric value indicating the CV threshold.
sort	logical constant indicating whether the filter MWAS_matrix will be sorted based on p-values.

Value

A numeric matrix corresponding to the filtered MWAS_matrix. The matrix has an additional column, which indicates the index of each metabolic variable in the original MWAS_matrix.

Examples

## Load data
data(targetMetabo_SE)

## Test for association between diabetes and target_metabolites
T2D_model <- MWAS_stats (targetMetabo_SE, disease_id = "T2D",
                         assoc_method = "logistic")

## Filter T2D_model by p-value
MWAS_filter(T2D_model, type = "pvalue", alpha_th = 0.001, sort = TRUE)

## Subset targetMetabo_SE based on pvalue_filter
pvalue_filter <- MWAS_filter(T2D_model, type = "pvalue", alpha_th = 0.001)
index_features <- pvalue_filter[, 4]
targetMetabo_SE[index_features, ]

MWASTools documentation built on Nov. 8, 2020, 5:07 p.m.