DesignModelMatrix: Create a Model Matrix for Experimental Design

In eisascience/scCustFx: single-cell custom fxs Package

Create a Model Matrix for Experimental Design

Description

This function generates a model matrix for statistical analysis based on metadata. It creates a grouping variable from the specified contrast columns, ensuring proper formatting for downstream analysis.

Usage

DesignModelMatrix(meta_df, contrast_columns, sampleIdCol = "cDNA_ID")

Arguments

meta_df	A data frame containing sample metadata.
contrast_columns	Character vector. Column names in 'meta_df' to be used for defining experimental groups.
sampleIdCol	Character. Name of the column in 'meta_df' that contains sample identifiers (default: '"cDNA_ID"').

Details

This function:

• Merges ('unites') values from 'contrast_columns' into a single 'group' column.

• Ensures valid R variable names using 'make.names()'.

• Constructs a design matrix using 'stats::model.matrix()', encoding the experimental conditions.

• Attaches metadata attributes ('contrast_columns', 'sampleIdCol') to the resulting matrix.

Value

A model matrix where rows correspond to samples and columns correspond to experimental conditions. The matrix includes attributes '"contrast_columns"' and '"sampleIdCol"' for reference.

Examples

## Not run: 
if(requireNamespace("dplyr", quietly = TRUE) && requireNamespace("tidyr", quietly = TRUE)) {
    metadata <- data.frame(
        cDNA_ID = c("Sample1", "Sample2", "Sample3"),
        Treatment = c("A", "B", "A"),
        Batch = c("X", "X", "Y")
    )
    design_matrix <- DesignModelMatrix(metadata, contrast_columns = c("Treatment", "Batch"))
    print(design_matrix)
}

## End(Not run)

eisascience/scCustFx documentation built on June 2, 2025, 3:59 a.m.