The iPQF spectra-to-protein summarisation method integrates peptide spectra characteristics and quantitative values for protein quantitation estimation. Spectra features, such as charge state, sequence length, identification score and others, contain valuable information concerning quantification accuracy. The iPQF algorithm assigns weights to spectra according to their overall feature reliability and computes a weighted mean to estimate protein quantities. See also combineFeatures for a more general overview of feature aggregation and examples.

iPQF(
object,
groupBy,
low.support.filter = FALSE,
ratio.calc = "sum",
method.combine = FALSE,
feature.weight = c(7, 6, 4, 3, 2, 1, 5)^2
)

object An instance of class MSnSet containing absolute ion intensities. Vector defining spectra to protein matching. Generally, this is a feature variable such as fData(object)$accession. A logical specifying if proteins being supported by only 1-2 peptide spectra should be filtered out. Default is FALSE. Either "none" (don't calculate any ratios), "sum" (default), or a specific channel (one of sampleNames(object)) defining how to calculate relative peptides intensities. A logical defining whether to further use median polish to combine features. Vector "numeric" giving weight to the different features. Default is the squared order of the features redundant -unique-distance metric, charge state, ion intensity, sequence length, identification score, modification state, and mass based on a robustness analysis. ## Value A matrix with estimated protein ratios. ## References iPQF: a new peptide-to-protein summarization method using peptide spectra characteristics to improve protein quantification. Fischer M, Renard BY. Bioinformatics. 2016 Apr 1;32(7):1040-7. doi:10.1093/bioinformatics/btv675. Epub 2015 Nov 20. PubMed PMID:26589272. ## Author Martina Fischer ## Examples data(msnset2) head(exprs(msnset2)) #> X114.ions X115.ions X116.ions X117.ions #> 11695 7.0384398 7.6349574 7.2886949 6.7348692 #> 11696 9.4098886 10.1034528 10.0697640 9.4273709 #> 11697 4.5859834 4.6926775 4.5016924 4.2924254 #> 11698 3.7262916 3.8913142 3.7928041 3.3721019 #> 11699 1.2229379 1.4005805 1.3940034 1.2957861 #> 11700 0.8303542 0.7397313 0.7686411 0.7402981 prot <- combineFeatures(msnset2, groupBy = fData(msnset2)$accession,
method = "iPQF")
#> The following 1 proteins are only supported by 1 or 2 peptide spectra,
#> hence, protein quantification is not reliable and can only be calculated
#> by the 'mean' in these cases, corresponding protein accessions are:
#>   O95678