dataFilter: Filter low variance genes by given minimal 'mad' value or...

In Tong-Chen/ImageGP: For ImageGP cloud platform (https://www.bic.ac.cn/BIC)

Filter low variance genes by given minimal mad value or keep top number/percent genes with bigger variances.

Description

Filter low variance genes by given minimal mad value or keep top number/percent genes with bigger variances.

Usage

dataFilter(
  datExpr,
  minimal_mad = NULL,
  top_mad_n = 0.75,
  rmVarZero = T,
  noLessThan = NULL,
  value_type = mad
)

Arguments

datExpr	Normal gene expression matrix (gene x sample).
minimal_mad	Minimal allowed mad value.
top_mad_n	An integer larger than 1 will be used to get top x genes (like top 5000). A float number less than 1 will be used to get top x fraction genes (like top 0.7 of all genes).
rmVarZero	Default TRUE. Remove genes with variance as 0. Normally for PCA or correlation analysis.
noLessThan	Specify the lowest number of genes to be kept. Default NULL meaning no lower limit.
value_type	Specify the way for statistical computation. Default mad, accept mean, var.

Value

A dataframe.

Examples

df = generateAbundanceDF(nSample=30, nGrp=3)
dataFilter(df)

Tong-Chen/ImageGP documentation built on April 14, 2025, 12:54 p.m.