Skip to content

Separate FDR estimation

Bo edited this page Apr 16, 2019 · 6 revisions

Separate FDR estimation

The algorithm we used to calculate separate FDR was described in the PGA paper.

Input PSM file

The input PSM file must be in TSV format and the following columns are required:

  • index: Spectrum ID
  • peptide: Peptide sequence
  • charge: Precursor charge
  • score: The score of PSM
  • protein: Multiple protein IDs are separated by ";"

Example

PSM level FDR estimation

library(PGA)
calculateFDR(psmfile="all_psm.txt",
             db="customized_db.fasta",
             fdr=0.01,
             decoyPrefix="XXX_",
             novelPrefix="VAR",
             better_score_lower=TRUE,
             remap=FALSE,
             out_dir="./",
             protein_inference=FALSE,
             xmx=4)

Set parameter protein_inference as TRUE to perform protein inference.

Peptide level FDR estimation

library(PGA)
calculateFDR(psmfile="all_psm.txt",
             db="customized_db.fasta",
             peptide_level=TRUE,
             fdr=0.01,
             decoyPrefix="XXX_",
             novelPrefix="VAR",
             better_score_lower=TRUE,
             remap=FALSE,
             out_dir="./",
             protein_inference=FALSE,
             xmx=4)
Clone this wiki locally