similarity_bootstrap: Bootstraps training and prediction of scRNA-seq data using...

In zktuong/kelvinny: Kelvinny The Incredible!

Bootstraps training and prediction of scRNA-seq data using binomial/multinomial logistic regression with tunable penalties

Description

Bootstraps training and prediction of scRNA-seq data using binomial/multinomial logistic regression with tunable penalties

Usage

similarity_bootstrap(
  trainingSet,
  trainingCellType,
  testingSet,
  nboots = 50,
  simplify = FALSE,
  percent_probability = TRUE,
  bs_nCores = 10,
  train_nCores = parallel::detectCores(),
  verbose = FALSE,
  ...
)

Arguments

trainingSet	Seurat object, SummarizedExperiment object or expression matrix for training
trainingCellType	The cell types/clusters in the training data set
testingSet	Seurat object, SummarizedExperiment object or expression matrix for testing
nboots	integet specifying number of bootstrap runs to run
simplify	logical. collapse the predictions to a single data frame
percent_probability	logical. return as percentage or logit
bs_nCores	integer specifying number of cores for parallelization (for bootstrap).
train_nCores	integer specifying number of cores for parallelization (for glmnet).
verbose	logical. if TRUE, print the training/prediction steps. Otherwise, minimal messages will be returned on the screen
...	other functions pass to glmnet and predict. see ?glmnet and ?predict.glmnet

Value

Generates prediction.

Examples

pred <- predScSimilarity(model, test.sce)

zktuong/kelvinny documentation built on July 18, 2024, 10:42 a.m.