createDesignMatrix: Create design matrix.

In lmweber/regsplice: L1-regularization based methods for detection of differential splicing

Create design matrix.

Description

Create a model design matrix for a single gene.

Usage

createDesignMatrix(condition, n_exons)

Arguments

condition	Experimental conditions for each sample (character or numeric vector, or factor).
n_exons	Number of exons in the gene (integer).

Details

Creates a model design matrix for a single gene in the format required by the regsplice model fitting functions. Required inputs are the experimental conditions (groups) for each sample, and the number of exons in the gene.

The design matrix includes main effect terms for each exon and each sample, and interaction terms between the exons and conditions.

Note that the design matrix does not include main effect terms for the conditions, since these are absorbed into the main effect terms for the samples. In addition, the design matrix does not include an intercept column, since it is simpler to let the model fitting functions add an intercept term later.

The model fitting functions in subsequent steps call this function once for each gene.

Value

Returns a model design matrix for the gene, in the format required by the regsplice model fitting functions.

See Also

fitRegMultiple fitNullMultiple fitFullMultiple LRTests

Examples

condition <- rep(c(0, 1), each = 3)
n_exons <- 10
X <- createDesignMatrix(condition, n_exons)

lmweber/regsplice documentation built on March 19, 2024, 1:45 p.m.