Skip to contents

Filter out unreliable PSMs.

Source: R/PSM-filter.R
filterPSMs.Rd

Functions to filter out PSMs matching. The PSMs should be stored in a PSM such as those produced by PSM().

  • filterPsmDecoy() filters out decoy PSMs, i.e. those annotated as isDecoy.

  • filterPsmRank() filters out PSMs of rank > 1.

  • filterPsmShared() filters out shared PSMs, i.e. those that match multiple proteins.

  • filterPsmFdr() filters out PSMs based on their FDR.

Usage

filterPSMs(
  x,
  decoy = psmVariables(x)["decoy"],
  rank = psmVariables(x)["rank"],
  protein = psmVariables(x)["protein"],
  spectrum = psmVariables(x)["spectrum"],
  peptide = psmVariables(x)["peptide"],
  verbose = TRUE
)

filterPsmDecoy(x, decoy = psmVariables(x)["decoy"], verbose = TRUE)

filterPsmRank(x, rank = psmVariables(x)["rank"], verbose = TRUE)

filterPsmShared(
  x,
  protein = psmVariables(x)["protein"],
  peptide = psmVariables(x)["peptide"],
  verbose = TRUE
)

filterPsmFdr(x, FDR = 0.05, fdr = psmVariables(x)["fdr"], verbose = TRUE)

Arguments

x

An instance of class PSM.

decoy

character(1) with the column name specifying whether entries match the decoy database or not. Default is the decoy PSM variable as defined in psmVariables(). The column should be a logical and only PSMs holding a FALSE are retained. Filtering is ignored if set to NULL or NA.

rank

character(1) with the column name holding the rank of the PSM. Default is the rank PSM variable as defined in psmVariables(). This column should be a numeric and only PSMs having rank equal to 1 are retained. Filtering is ignored if set to NULL or NA.

protein

character(1) with the column name holding the protein (groups) protein. Default is the protein PSM variable as defined in psmVariables(). Filtering is ignored if set to NULL or NA.

spectrum

character(1) with the name of the spectrum identifier column. Default is the spectrum PSM variable as defined in psmVariables(). Filtering is ignored if set to NULL or NA.

peptide

character(1) with the name of the peptide identifier column. Default is the peptide PSM variable as defined in psmVariables(). Filtering is ignored if set to NULL or NA.

verbose

logical(1) setting the verbosity flag.

FDR

numeric(1) to be used to filter based on the fdr variable. Default is 0.05.

fdr

character(1) variable name that defines that defines the spectrum FDR (or any similar/relevant metric that can be used for filtering). This value isn't set by default as it depends on the search engine and application. Default is NA.

Value

A new filtered PSM object with the same columns as the input x.

Author

Laurent Gatto

Examples

f <- MsDataHub::TMT_Erwinia_1uLSike_Top10HCD_isol2_45stepped_60min_01.20141210.mzid()
#> see ?MsDataHub and browseVignettes('MsDataHub') for documentation
#> loading from cache

id <- PSM(f)
filterPSMs(id)
#> Starting with 5802 PSMs:
#> Removed 2896 decoy hits.
#> Removed 155 PSMs with rank > 1.
#> Removed 85 shared peptides.
#> 2666 PSMs left.
#> PSM with 2666 rows and 35 columns.
#> names(35): sequence spectrumID ... subReplacementResidue subLocation