vstran: Variance stablization transformation

In MonkeyLB/hicrep: Measuring the reproducibility of Hi-C data

Description

The function finds the rank of the data and rescale the ranks in the range of (0, 1).

Usage

vstran(d)

Arguments

d	a matrix or data.frame that has two columns, each column is the Hi-C read counts data in a replicate.

Details

In Hi-C data, the read counts for contacts with short interaction distances have a much larger dynamic range than those with long interaction distances. To mitigate this difference, we rank the contact counts in each stratum separately, and then normalize the ranks by the total number of observations in each stratum, such that all strata share a similar dynamic range.

Value

a matrix of two columns, each represents a transformed read counts of a replicate.

Examples

data(HiCR1)
data(HiCR2)

#exact contact counts within 1Mb interaction distance
d1 <- d2 <- NULL
for (i in 1:nrow(HiCR1)){
    d1 = c(d1, HiCR1[i, i+1])
    d2 = c(d2, HiCR2[i, i+1])
    d = cbind(d1, d2)
}

HiC_R1_vs <- vstran(d)
head(HiC_R1_vs)

MonkeyLB/hicrep documentation built on Dec. 15, 2020, 12:47 a.m.