Skip to contents

Normalize proteomics data

Source: R/Normalization.R
normalize_proteomics.Rd

Complete normalization pipeline: zero-to-NA conversion, protein filtering by group presence, SummarizedExperiment creation, and normalization with the selected method. Output is always in log2 scale.

Usage

normalize_proteomics(
  data,
  metadata = NULL,
  min_reps = NULL,
  min_groups = 1,
  norm_method = "cycloess",
  cyclic_loess_method = c("fast", "pairs"),
  cyclic_loess_iterations = 3,
  cyclic_loess_span = 0.7,
  verbose = TRUE
)

Arguments

data

Either a proteomics_data object – the output of preprocess_spectronaut(), preprocess_diann(), preprocess_tmt() or preprocess_lfq() – in which case metadata is derived from it and must be left NULL, or a data frame with ProteinGroups, GeneNames, UniqPepts and the intensity columns.

metadata

Data frame with Column, Condition, Replicate. Only needed when data is a data frame.

min_reps

Minimum replicates for filtering. NULL = auto: floor(min_group_size / 2)

min_groups

Minimum groups meeting min_reps (default: 1)

norm_method

Normalization method (default: "cycloess"). One of:

  • Group A (input: raw intensities): "log2Norm", "GlobalMedian", "GlobalMean", "eqmedians", "vsn"

  • Group B (input: log2 assay): "log2" (no extra normalization), "quantile", "Rlr", "MAD", "cycloess", "medianNorm", "meanNorm", "quantile.robust"

cyclic_loess_method

Cyclic Loess method: "fast" or "pairs" (default: "fast")

cyclic_loess_iterations

Number of iterations for Cyclic Loess (default: 3)

cyclic_loess_span

Span parameter for Cyclic Loess (default: 0.7)

verbose

Print progress messages (default: TRUE)

Value

List with:

  • se: SummarizedExperiment with assays raw, log2, and <norm_method> (assays raw + log2 only when norm_method = "log2")

  • filter_summary: Filtering summary

  • na_overview: NA statistics

Examples

data(nadia_dia)

# The object returned by preprocess_*() is accepted directly
norm <- normalize_proteomics(nadia_dia, norm_method = "cycloess",
                             verbose = FALSE)
SummarizedExperiment::assayNames(norm$se)
#> [1] "raw"      "log2"     "cycloess"

# Zeros become NA and low-coverage proteins are filtered out
norm$filter_summary
#> $n_total
#> [1] 2000
#> 
#> $n_keep
#> [1] 1997
#> 
#> $n_drop
#> [1] 3
#> 
#> $min_reps
#> [1] 2
#> 
#> $min_groups
#> [1] 1
#> 

# Any of the 13 methods can be used
norm2 <- normalize_proteomics(nadia_dia, norm_method = "quantile",
                              verbose = FALSE)
SummarizedExperiment::assayNames(norm2$se)
#> [1] "raw"      "log2"     "quantile"