Read glyco-decipher output

Glyco-Decipher is a software for glycopeptide identification and quantification. This function reads in the result file and returns a glyexp::experiment() object. Currently only label-free quantification is supported.

Usage

read_glyco_decipher(
  fp,
  sample_info = NULL,
  quant_method = "label-free",
  glycan_type = "N",
  sample_name_converter = NULL,
  orgdb = "org.Hs.eg.db"
)

Arguments

fp

File path of the pGlyco3 result file.

sample_info

File path of the sample information file (csv), or a sample information data.frame/tibble.

quant_method

Quantification method. Either "label-free" or "TMT".

glycan_type

Glycan type. Either "N" or "O". Default is "N".

sample_name_converter

A function to convert sample names from file paths. The function should take a character vector of old sample names and return new sample names. Note that sample names in sample_info should match the new names. If NULL, original names are kept.

orgdb

name of the OrgDb package to use for UniProt to gene symbol conversion. Default is "org.Hs.eg.db".

Value

An glyexp::experiment() object.

Which file to use?

You should use the "site.csv" file in the result folder. This file contains "Site" and "Glycan" columns, followed by quantification result for each sample.

Protein inference and uncertain sites

Glyco-Decipher reports uncertain sites and proteins. This function automatically performs protein inference using the parsimony method to find the leader proteins. This ensures each glycopeptide is uniquely mapped to a single protein, gene, and glycosite. Besides, uncertain sites on proteins are assigned NA.

Variable information

The following columns could be found in the variable information tibble:

• protein: character, protein accession (after protein inference)

• protein_site: integer, site of glycosylation on protein (after protein inference)

• gene: character, gene name (symbol) (after protein inference)

• glycan_composition: glyrepr::glycan_composition(), glycan compositions.

Sample information

The sample information file should be a csv file with the first column named sample, and the rest of the columns being sample information. The sample column must match the RawName column in the pGlyco3 result file, although the order can be different.

You can put any useful information in the sample information file. Recommended columns are:

• group: grouping or conditions, e.g. "control" or "tumor", required for most downstream analyses

• batch: batch information, required for batch effect correction