QuantNorm (version 1.0.5)

QuantNorm: Adjust the distance matrix by quantile normalization for data with batch effect

Description

This function applies quantile normalization on the distance matrix (dissimilarity matrix) and return the corrected distance matrix.

Usage

QuantNorm(dat, batch, method = "row/column", cor_method = "spearman",
  tol = 0.01, max = 50, logdat = TRUE, standardize = FALSE)

Arguments

dat

The original p*n batch effect data with n subjects and p RNA-seq measurements or the n by n distance matrix.

batch

The vector of length n indicating which batch the subjects belong to.

method

Method for the quantile normalization. There are two options: "row/column" and "vectorize".

cor_method

Method to calculate the correlation matrix, can be 'spearman'(default), 'pearson' or 'kendall'.

tol

The tolerance for the iterative method "row/column", which is the Euclidean distance of the vectorized two dissimilarity matrices before and after each iteration.

max

Maximum number of the iteration if the tolerance is not reached.

logdat

Whether conducting log transformation to data or not.

standardize

Whether conducting standardization [(dat - mean)/sqrt(var)] to data or not.

Value

Returns the corrected 1-correlation matrix between subjects.

References

Fei et al (2018), Mitigating the adverse impact of batch effects in sample pattern detection, Bioinformatics, <https://doi.org/10.1093/bioinformatics/bty117>.

Examples

# NOT RUN {
library(pheatmap) #drawing heatmap

data("ENCODE") #load the ENCODE data

#Before correction, the subjects are clustered by species
pheatmap(cor(ENCODE))

#Assigning the batches based on species
batches <- c(rep(1,13),rep(2,13))

#QuantNorm correction
corrected.distance.matrix <- QuantNorm(ENCODE,batches,method='row/column', cor_method='pearson',
                                       logdat=FALSE, standardize = TRUE, tol=1e-4)
pheatmap(1-corrected.distance.matrix)
# }