Input File Description

Petra Palenikova & Rick Scavetta

2022-04-26

Peptide input file

system.file(“extData”, “data.txt”, package = “ComPrAn”) file in the example data set
Input file needs to be a text file with tab deliminated values. Data are organized in a column format, first row contain column headers. This file must contain following columns, format of values inside each column is also specified:

Description of columns

“PSM Ambiguity” - rows in which “PSM Ambiguity” value is “Rejected” will be excluded from analysis

“Precursor Area” - column containing quantification of peptides

“# Protein Groups” - specifies to how many protein groups a specific peptide was assigned

“Rank” - Rank assigned to the peptide by the upstream programms used

“Confidence Level” - Confidence assigned to the peptide by the upstream programms used

“Protein Group Accessions” - UniProt IDs of proteins

“Protein Descriptions” - description of the protein, for proper functionality of plotting functions should contain “GN=” with gene name after the equal sign

“Modifications” - chemical modifications of the peptide, including labeling by heavy amino acids, if a peptide has multiple modifications values should be separated by semicolon followed by space “;”, heavy amino acids should be noted in a format containing word “Label” in its value

“Charge” - charge state of peptide

“Sequence” - amino acid sequence of peptide in single letter format

“Search ID” - column specifiyng fraction in a format where “A” means fraction 1, “B” fraction 2 and so on, for samples with more then 26 fractions the values continue as “AA” for 27, “AB” for 28 etc. By default, the name of this columns is “Search ID” but a different name can be specified when calling ‘cleanData()’ function

Fig 1 Example minimal peptide input file

Fig 1 Example minimal peptide input file

Example peptide file was produced using Proteome Discoverer™ Software (Thermo Scientific™)

Normalised proteins input file

dataNormProts.txt file in an example data set
Input file needs to be a text file with tab deliminated values. Data are organized in a column format, first row contain column headers. This file must contain following columns, format of values inside each column is also specified:

Description of columns

“Protein Group Accessions” - UniProt IDs of proteins

“Protein Descriptions” - description of the protein, for proper functionality of plotting functions should contain “GN=” with gene name after the equal sign

“scenario” - A or B, specifies whether the values in any given row belog to “scenario A” or “scenario B”, meaning of these two scenarios is described in the SILACcomplexomics vignette

“label” - specifies whether the values in any given row belog to “labeled” (=TRUE) or “unlabeled”(=FALSE) sample

columns “1” to “n” - n represents total number of fractions in sample, each row contains normalised intensity values for a given protein in a given fraction

Fig 2 Example normalised protein input file

Fig 2 Example normalised protein input file