scLearn_model_learning: Training the learning-based model with scLearn method.

In bm2-lab/scLearn: Learning for single cell assignment

Training the learning-based model with scLearn method.

Description

Training the learning-based model with scLearn method.

Usage

scLearn_model_learning(high_varGene_names, expression_profile, sample_information_cellType,sample_information_timePoint=NULL, bootstrap_times = 10, cutoff = 0.01,dim_para=0.999)

Arguments

high_varGene_names	A vecter showing the selected genes.
expression_profile	A dataframe showing the reference expression profile. The row is gene and The column is sample.
sample_information_cellType	A character vector showing the cell type of each sample. The column name of the vector is the sample name.
sample_information_timePoint	A character vector showing the time point of each sample. The column name of the vector is the sample name. The default is NULL.
bootstrap_times	The times for bootstrapping which should be at least larger than 1. Default is 10.
cutoff	The cutoff for selecting similarity threshold for each cell type. Default is 0.01.
dim_para	The threshold to choose proper dimension for MDDM. Default is 0.999.

Value

high_varGene_names	A vecter showing the selected genes.
simi_threshold_learned	A list showing the similarity threshold for each bootstrapping when "sample_information_timePoint" is NULL, or it is a vector showing the similarity threshold.
feature_matrix_learned	A list showing the learned feature for each bootstrapping when "sample_information_timePoint" is NULL, or it is a matrix showing the learned feature.
trans_matrix_learned	A list showing the learned transformation matrix for each bootstrapping when "sample_information_timePoint" is NULL, or it is a matrix showing the learned transformation matrix.

Author(s)

Bin Duan

bm2-lab/scLearn documentation built on Dec. 17, 2024, 8:18 p.m.