noiseBaseFit {RaceID}

Function for computing a fit to the baseline of gene expression variability

Description

This function fits a second order polynomial to the baseline variance-mean dependence across all genes in log space.

Usage

noiseBaseFit(x, step = 0.01, thr = 0.05)

Arguments

x

Matrix of gene expression values with genes as rows and cells as columns.

step

Positive real number between 0 and 1. Bin size for the computation. The interval of mean gene expression values is divided into bins with equal number of data points and step equals the fraction of data points in each bin. Default is 0.01.

thr

Positive real number between 0 and 1. In each mean expression bin defined by step the lowest thr-quantile of the gene expression variance distribution is selected. The selected data points from all bins are used for a second order polynomial fit of the variance-mean dependence in log space. Default is 0.05.

Value

List object of three components:

nfit

model fit as returned by the lm function.

m

mean expression of all genes

v

expression variance of all genes

Examples

x <- noiseBaseFit(intestinalDataSmall,step=.01,thr=.05)