findClassyGenes: Find Classifier-Worthy Gene Candidates

In pcahan1/singleCellNet: cell identity from single cell RNA-Seq data

Description

Find classifier-worthy genes for each cancer category to train the classifier

Usage

findClassyGenes(expDat, sampTab, dLevel, topX = 25, dThresh = 0,
  alpha1 = 0.05, alpha2 = 0.001, mu = 2)

Arguments

expDat	a matrix of normalized expression data from trans_prop
sampTab	a dataframe of the sample table
dLevel	a string indicating the column name in sample table that contains the cancer category
topX	an integer indicating the number of top positive classification genes for each category to select for training. Will also select topX number of negative classification genes.
dThresh	a number representing the detection threshold
alpha1	a number representing proportion of cells in which a gene must be considered detected (as defined in geneStats)
alpha2	a number representing lower proportion of cells for genes that must have higher expression level
mu	a number represeting threshold for average expression level of genes passing the lower proportion criteria

Value

a list containing two lists: a list of classifier worthy genes named 'cgenes' and a list of cancer category named 'grps'

pcahan1/singleCellNet documentation built on April 9, 2021, 8:49 a.m.