make_input: make_input

In hemberg-lab/sctransfermap: sctransfermap for predicting cell type in single-cell data (RNA-seq and ATAC-seq)

Description

make input format for sctransfer map from matrix file

Usage

make_input(input_var, min_thr = 0.1, max_thr = 0.9, pseudo_cnt = 1,
  train = FALSE, gene_usage = FALSE)

Arguments

input_var	matrix class composed of gene/peak_location(row)-by-cell(column) - same as result of importdata or convert_seurat
min_thr	minimum threshold of expressing cell ratio for a given gene/peak_location
max_thr	maximum threshold of expressing cell ratio for a given gene/peak_location
pseudo_cnt	pseudo count for count matrix for log-transformation
train	TRUE if it is data for training TRUE: adjust min_thr and max_thr. In addition, make the gene/peak_location count for hadamard matrix FALSE: no gene/peak_location filtering
gene_usage	option for gene/peak_location filtering to make hadamard matrix FALSE: gene/peak_location filtering by discarding low expressing cell ratio genes/peak_locations certain number (ex: 2048): the number of gene to use for training (must be suitable number to generate hadamard matrix)

Value

return a list variable composed of 4 objects

• norm_matrix: expression or peak matrix (row: gene/peak_location, col: cell)

• rowname: gene or peak location

• cellname: cell annotation

• celltype: celltype == sort(unique(cellname))

Examples

data1 <- make_input(input_var) [query case]
data1 <- make_input(input_var, train = TRUE) [training case]

hemberg-lab/sctransfermap documentation built on Nov. 11, 2019, 6:23 a.m.