Chris, xinteract chooses just one single way of running PTMProphet through the method you showed. While PTMProphet.cxx is still in the repository it is not active code. The PTMProphet used class is called PTMProphetMpx and it doesn't rely on the decoys. Sorry for the confusion.
-David On Wed, Jun 4, 2014 at 10:50 AM, Chris <[email protected]> wrote: > I'll show you my logic then just in case it is in an actual functioning bit > of code: > > From xinteract (which TPP launches to handle everything it seems): > > -i[iProphet options] [run iProphet on the PeptideProphet > result] > iProphet options [following the 'i']: > .... > -M[PTMProphet options] [run PTMProphet on the iProphet > result] > PTMProphet options [following the 'M' (e.g. -M-STY,79.9663-MZTOL=0.4)]: > -{<amino acids, n, or c>,<mass_shift>,...} [specify mod > masses (e.g. -STY,79.9663,K,114.0429,M,15.9949)] > -MZTOL=<number> [Use specified > +/- mz tolerance on site specific ions (default=0.1 dalton)] > -NOUPDATE [Don't > update modification_info tags in pepXML] > > which indicated to me the workflow is: > > Pep Proph->iProph->PTMProphet > > The code from PTMProphet which makes me believe it needs incorrect hits: > PTMProphet.cxx - line 216 comments on using decoys for the null > distribution, and is referenced in the for loop for line 391 and onward. > > As a reference I'm using Polar Vortex > > On Wednesday, June 4, 2014 1:30:18 PM UTC-4, David Shteynberg wrote: >> >> Chris, >> >> You must be looking at a non-functioning section of the code. PTMProphet >> relies only on the MS/MS spectrum information to determine the PTM site >> scores. PTMProphet assumes the identification is the correct peptide and >> attempts to find the most likely sites for the PTMs. It will work on any >> pepXML input file if run on the commandline, so either way of using the tool >> is fine. >> >> Cheers, >> -David >> >> >> On Wed, Jun 4, 2014 at 6:42 AM, Chris <[email protected]> wrote: >>> >>> Hi everyone, >>> >>> I'm a bit confused about the workflow for finding PTMs. At first glance >>> and seeing the code for xinteract, it seems it should be: >>> >>> Search->PeptideProphet->InterProphet (apply 1% FDR)->PTMProphet >>> >>> However, looking at the code for PTMProphet, it uses the decoy hits to >>> build its null distribution. So my intuition says the following workflow >>> would be correct: >>> >>> Search->PeptideProphet->PTMProphet->InterProphet (apply 1% FDR) >>> >>> Because we want to have the false hits to build the null distribution in >>> PTMProphet. >>> >>> Can anyone clarify these points? >>> >>> Thank you, >>> Chris >>> >>> -- >>> You received this message because you are subscribed to the Google Groups >>> "spctools-discuss" group. >>> To unsubscribe from this group and stop receiving emails from it, send an >>> email to [email protected]. >>> To post to this group, send email to [email protected]. >>> >>> Visit this group at http://groups.google.com/group/spctools-discuss. >>> For more options, visit https://groups.google.com/d/optout. >> >> > -- > You received this message because you are subscribed to the Google Groups > "spctools-discuss" group. > To unsubscribe from this group and stop receiving emails from it, send an > email to [email protected]. > To post to this group, send email to [email protected]. > Visit this group at http://groups.google.com/group/spctools-discuss. > For more options, visit https://groups.google.com/d/optout. -- You received this message because you are subscribed to the Google Groups "spctools-discuss" group. To unsubscribe from this group and stop receiving emails from it, send an email to [email protected]. To post to this group, send email to [email protected]. Visit this group at http://groups.google.com/group/spctools-discuss. For more options, visit https://groups.google.com/d/optout.
