dot-fitGLM: .fitGLM This function parallelizes glm models: Parallelizes...

In gofflab/glmpd: Identification of context-specific drivers of a pattern projection via GLM fitting.

Description

.fitGLM This function parallelizes glm models: Parallelizes across genes in a counts matrix Parallelizes across patterns in a matrix of projected patterns

Usage

.fitGLM(
  countsMatrix,
  annotDF,
  model_formula_str,
  projected_patterns,
  cores,
  geneDF = NULL
)

Arguments

countsMatrix	matrix of gene expression counts. genes x cells / samples.
annotDF	A dataframe with annotations for each cell or sample in the gene expression dataset
model_formula_str	A string specifying the model to fit, in the format "~ var1 + var2"
projected_patterns	A matrix of projected pattern weights to use as an explanatory variable in the model fit for each gene. colnames are patterns, rownames are cells. Must have viable column names.
cores	Integer defining the number of cores to use for parallelization of model fitting across genes.
geneDF	(currently) optional dataframe with annotations for each gene in the dataset
exp_family	The regression family to use (default: 'negbinomial')
clean_model	Boolean. Passed through to monocle3::fit_model
verbose	Boolean. Verbose output for model fitting.
result	Return full or summarized results. Summarized results are stripped of memory-intensive models and return coefficients from extractCoefficients().

Value

Returns a list of fitted models (output similar to monocle3::fit_model)

gofflab/glmpd documentation built on April 11, 2021, 6:38 a.m.