Skip to contents

Quantile Normalization

Source: R/normalization_quantile.R
quantile_normalization.Rd

Normalizes read counts by the quantile normalization method:

  1. Each sample (column) is sorted, and values at each rank are averaged across columns

  2. Each sample's values are replaced with the average of their respective rank

  3. If log_trans = TRUE, applies log2(QN + 1) transformation

Usage

quantile_normalization(
  x,
  log_trans = FALSE,
  assay_name = NULL,
  new_assay_name = NULL
)

Arguments

x

A numeric matrix or data.frame of gene counts, or a SummarizedExperiment containing such counts.

If a SummarizedExperiment,

the function applies normalization to the specified assay (via assay_name).

If a data.frame/matrix,

the normalization is applied directly.

log_trans

Logical. If TRUE, apply log2(... + 1) transform to the quantile-normalized values.

assay_name

If x is a SummarizedExperiment, name of the assay to normalize. Defaults to the first assay if not specified.

new_assay_name

If x is a SummarizedExperiment, name of a new assay in which to store the quantile-normalized (or log2-transformed) values. If NULL, overwrites the original assay.

Value

A numeric matrix of quantile-normalized (or log2-normalized) values if x is a data.frame or matrix. If x is a SummarizedExperiment, returns the modified SummarizedExperiment with the normalized data placed in the existing or new assay.

Details

If x is a SummarizedExperiment, the function will extract the assay using assay_name, apply quantile normalization, and return a new or updated assay. If x is a matrix or data.frame, normalization is applied directly to the input matrix.

Examples

library(SummarizedExperiment)
library(airway)
data('airway')

se <- airway

# Only use a random subset of 1000 rows
set.seed(123)
idx <- sample(seq_len(nrow(se)), size = min(1000, nrow(se)))
se <- se[idx, ]

# -------------------------------
# 1) Using a data.frame
# -------------------------------
df <- assay(se)

## Without log transformation
df_qn <- quantile_normalization(df, log_trans = FALSE)
df_qn[1:5, 1:5]
#>                 SRR1039508 SRR1039509 SRR1039512 SRR1039513 SRR1039516
#> ENSG00000260166      0.000      0.000      0.000        0.0       0.00
#> ENSG00000266931      0.000      0.000      0.000        0.0       0.00
#> ENSG00000104774   2164.875   2319.375   1946.875     1971.5    1610.75
#> ENSG00000267583      0.000      0.750      0.000        0.0       0.00
#> ENSG00000227581      0.625      0.000      0.000        0.0       0.00

## With log transformation
df_qn_log <- quantile_normalization(df, log_trans = TRUE)
df_qn_log[1:5, 1:5]
#>                 SRR1039508 SRR1039509 SRR1039512 SRR1039513 SRR1039516
#> ENSG00000260166  0.0000000  0.0000000    0.00000    0.00000    0.00000
#> ENSG00000266931  0.0000000  0.0000000    0.00000    0.00000    0.00000
#> ENSG00000104774 11.0807343 11.1801423   10.92769   10.94581   10.65441
#> ENSG00000267583  0.0000000  0.8073549    0.00000    0.00000    0.00000
#> ENSG00000227581  0.7004397  0.0000000    0.00000    0.00000    0.00000

# -------------------------------
# 2) Using a SummarizedExperiment
# -------------------------------

## Overwrite existing assay
se2 <- quantile_normalization(se)
assay(se2)[1:5, 1:5]
#>                 SRR1039508 SRR1039509 SRR1039512 SRR1039513 SRR1039516
#> ENSG00000260166      0.000      0.000      0.000        0.0       0.00
#> ENSG00000266931      0.000      0.000      0.000        0.0       0.00
#> ENSG00000104774   2164.875   2319.375   1946.875     1971.5    1610.75
#> ENSG00000267583      0.000      0.750      0.000        0.0       0.00
#> ENSG00000227581      0.625      0.000      0.000        0.0       0.00

## Store result in new assay
se3 <- quantile_normalization(se, new_assay_name = "quant_norm")
assay(se3, "quant_norm")[1:5, 1:5]
#>                 SRR1039508 SRR1039509 SRR1039512 SRR1039513 SRR1039516
#> ENSG00000260166      0.000      0.000      0.000        0.0       0.00
#> ENSG00000266931      0.000      0.000      0.000        0.0       0.00
#> ENSG00000104774   2164.875   2319.375   1946.875     1971.5    1610.75
#> ENSG00000267583      0.000      0.750      0.000        0.0       0.00
#> ENSG00000227581      0.625      0.000      0.000        0.0       0.00

## Use specific input assay (simulate new one)
new_matrix <- matrix(
  data = sample(x = seq(1, 100000), size = nrow(se) * ncol(se),
  replace = TRUE),
  nrow = nrow(se),
  ncol = ncol(se)
)
rownames(new_matrix) <- rownames(se)
colnames(new_matrix) <- colnames(se)

## Create a new assay in the SummarizedExperiment
assay(se, "new_counts") <- new_matrix

## Normalize the new assay and store it under a new name
se4 <- quantile_normalization(se, assay_name = "new_counts",
new_assay_name = "quant_new")
assay(se4, "quant_new")[1:5, 1:5]
#>                 SRR1039508 SRR1039509 SRR1039512 SRR1039513 SRR1039516
#> ENSG00000260166   56060.00   94285.88   82647.62   60109.62  32567.250
#> ENSG00000266931   54882.62   76292.25   79960.25   90485.50   3259.625
#> ENSG00000104774     733.25   40996.50    9360.25   58418.00  91851.750
#> ENSG00000267583   19468.50   65204.88   51248.62   78804.12  24111.875
#> ENSG00000227581   61854.62   21166.88   95707.25   40996.50  98769.375