Merges final peptide files

This function combines a list of peptide final peptide files into one single master file that is obtained by merging the unique peptides from the filtered original peptide files.
Additionally it can combine multiple final fragment files into a fragment library.

makeMaster(pepfiles, fragmentfiles, fdr = 0.01, method = c("BH",
  "Bonferroni", "qval"), span.rt = 0.05, span.int = 0.05,
  maxDeltaRt = Inf, removeNeutralLoss = TRUE, removePrecursor = TRUE,
  tolerance = 2.5e-05, verbose = interactive())

Arguments

pepfiles	A character vector of final peptide file names to be merged.
fragmentfiles	A character vector of final fragment file names to be combined into an fragment library. These files should be from the same runs as the final peptide files used in pepfiles.
fdr	A numeric indicating the peptide false discovery rate limit.
method	A character indicating the p-value adjustment to be used. One of BH (default), Bonferroni or qval.
span.rt	A numeric with the loess span parameter value to be used for retention time modelling.
span.int	A numeric with the loess span parameter value to be used for intensity modelling.
maxDeltaRt	A double value that sets a maximum limit for the retention time deviaton between master and slave run to be included
removeNeutralLoss	A logical, if TRUE peptides with neutral loss are removed from the fragment library.
removePrecursor	A logical, if TRUE precursor ions are removed from the fragment spectra.
tolerance	A double value that determines the tolerance used to look for the precursor ions.
verbose	A logical indicating if information should be printed out.

Value

An instance of class "MasterPeptides".

Details

The merging process is as follows:

1. Each individual peptide final peptide file is filtered to retain (i) non-duplicated unique tryptic peptides, (ii) peptides with a false discovery rate <= fdr and (iii) proteins with a false positive rate <= fpr.

2. The filtered peptide files are ordered (1) according to their total number of peptides (for example [P1, P2, P3]) and (2) as before with the first item is positioned last ([P2, P3, P1] in the previous example). The peptide data are then combined in pairs in these respective orders. The first one is called the master file.

3. For each (master, slave) pair, the slave peptide file retention times are modelled according to the (original) master's retention times and slave peptides, not yet present in the master file are added to the master file.

4. The final master datasets, containing their own peptides and the respective slave specific retention time adjusted peptides are returned as a MasterPeptides instance.

The resulting MasterPeptides instance can be further used for a complete master vs. peptides/Pep3D analysis, as described in Synapter, synergise or using the GUI (synapterGUI). To do so, it must be serialised (using the saveRDS function) with a .rds file extension, to be recognised (and loaded) as a R object.

When several quantitation (or identification) files are combined as a master set to be mapped back against the inidividual final peptide files, the second master [P2, P3, P1] is used when analysing the peptide data that was first selected in the master generation (P1 above). This is to avoid aligning two identical sets of peptides (those of P1) and thus not being able to generate a valid retention time model. This is detected automatically for the user.

The two master peptides dataframes can be exported to disk as two csv files with writeMasterPeptides. The MasterPeptides object returned by makeMaster can be saved to disk (with save or saveRDS) and later reloaded (with load or readRDS) for further analysis.

The fragment library generation works as follows:

1. Each individual final fragment file is imported and only peptides present in the master dataset are used.

2. The fragments are combined based on their precursor ions.

3. The intensities of identical fragments (seen in different runs) is summed and divided by the summed precursor intensity (of the same peptide in different runs).

4. Afterwards the intensities are normalized to the average precursor intensity of the different runs.

5. Finally a MSnExp object is created.

The fragment library dataframe can be exported to disk as csv file with writeFragmentLibrary.

References

Shliaha P.V., Bond N. J., Lilley K.S. and Gatto L., in prep.

See also

See the Synapter class manual page for detailed information on filtering and modelling and the general algorithm implemented in the synapter package.

The estimateMasterFdr function allows to control false dicovery rate when combining several peptide files while maximising the number of identifications and suggest which combination of peptide files to use.

The vignette, accessible with synapterGuide() illustrates a complete pipeline using estimateMasterFdr and makeMaster.